Remove random default argument value in the middle of argument list
[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 *
c997cbe6 9 * @copyright 1999-2021 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 $replace = preg_replace_callback('/=([0-9a-f]{2})/i',
906 create_function ('$matches', 'return chr(hexdec($matches[1]));'),
8bd0068d 907 $replace);
908 if ($utfencode) {
909 if ($can_be_encoded) {
910 /* convert string to different charset,
911 * if functions asks for it (usually in compose)
912 */
bcae324b 913 $replace = charset_convert($res[2], $replace,$default_charset,$htmlsafe);
8bd0068d 914 } else {
915 // convert string to html codes in order to display it
916 $replace = charset_decode($res[2], $replace);
917 }
918 } else {
bcae324b 919 if ($htmlsafe) {
3047e291 920 $replace = sm_encode_html_special_chars($replace);
8bd0068d 921 }
098ea084 922 }
8bd0068d 923 $ret .= $replace;
924 break;
925 default:
926 break;
79e07c7e 927 }
098ea084 928 $chunk = $res[5];
929 $encoded = true;
08b7f7cc 930 }
931 if (!$encoded) {
bcae324b 932 if ($htmlsafe) {
9f7f68c3 933 $ret .= '&#32;';
08b7f7cc 934 } else {
935 $ret .= ' ';
da2415c1 936 }
08b7f7cc 937 }
dc3d13a7 938
bcae324b 939 if (!$encoded && $htmlsafe) {
3047e291 940 $ret .= sm_encode_html_special_chars($chunk);
dc3d13a7 941 } else {
942 $ret .= $chunk;
943 }
098ea084 944 ++$i;
945 }
fd81e884 946 /* remove the first added space */
947 if ($ret) {
bcae324b 948 if ($htmlsafe) {
9f7f68c3 949 $ret = substr($ret,5);
fd81e884 950 } else {
951 $ret = substr($ret,1);
952 }
953 }
da2415c1 954
08b7f7cc 955 return $ret;
451f74a2 956}
957
9f7f68c3 958/**
a24cf710 959 * Encodes header
8bd0068d 960 *
a24cf710 961 * Function uses XTRA_CODE _encodeheader function, if such function exists.
a24cf710 962 *
758a7889 963 * Function uses Q encoding by default and encodes a string according to RFC
964 * 1522 for use in headers if it contains 8-bit characters or anything that
a24cf710 965 * looks like it should be encoded.
8bd0068d 966 *
758a7889 967 * Function switches to B encoding and encodeHeaderBase64() function, if
968 * string is 8bit and multibyte character set supported by mbstring extension
969 * is used. It can cause E_USER_NOTICE errors, if interface is used with
f270a6eb 970 * multibyte character set unsupported by mbstring extension.
971 *
8bd0068d 972 * @param string $string header string, that has to be encoded
973 * @return string quoted-printable encoded string
f270a6eb 974 * @todo make $mb_charsets system wide constant
8bd0068d 975 */
451f74a2 976function encodeHeader ($string) {
6fbd125b 977 global $default_charset, $languages, $squirrelmail_language;
83be314a 978
10dec454 979 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 980 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
33a55f5a 981 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
83be314a 982 }
793cc001 983
a24cf710 984 // Use B encoding for multibyte charsets
f270a6eb 985 $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
986 if (in_array($default_charset,$mb_charsets) &&
987 in_array($default_charset,sq_mb_list_encodings()) &&
988 sq_is8bit($string)) {
989 return encodeHeaderBase64($string,$default_charset);
990 } elseif (in_array($default_charset,$mb_charsets) &&
991 sq_is8bit($string) &&
992 ! in_array($default_charset,sq_mb_list_encodings())) {
993 // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
994 // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
995 }
a24cf710 996
451f74a2 997 // Encode only if the string contains 8-bit characters or =?
3d8371be 998 $j = strlen($string);
098ea084 999 $max_l = 75 - strlen($default_charset) - 7;
1000 $aRet = array();
451f74a2 1001 $ret = '';
c96c32f4 1002 $iEncStart = $enc_init = false;
0d53f0f9 1003 $cur_l = $iOffset = 0;
3d8371be 1004 for($i = 0; $i < $j; ++$i) {
1710ad65 1005 switch($string[$i])
c96c32f4 1006 {
6b76cffa 1007 case '"':
8bd0068d 1008 case '=':
1009 case '<':
1010 case '>':
1011 case ',':
1012 case '?':
1013 case '_':
1014 if ($iEncStart === false) {
1015 $iEncStart = $i;
1016 }
1017 $cur_l+=3;
1018 if ($cur_l > ($max_l-2)) {
1019 /* if there is an stringpart that doesn't need encoding, add it */
08b7f7cc 1020 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
c96c32f4 1021 $aRet[] = "=?$default_charset?Q?$ret?=";
1022 $iOffset = $i;
1023 $cur_l = 0;
1024 $ret = '';
1025 $iEncStart = false;
08b7f7cc 1026 } else {
1710ad65 1027 $ret .= sprintf("=%02X",ord($string[$i]));
c96c32f4 1028 }
8bd0068d 1029 break;
1030 case '(':
1031 case ')':
1032 if ($iEncStart !== false) {
08b7f7cc 1033 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
8bd0068d 1034 $aRet[] = "=?$default_charset?Q?$ret?=";
c96c32f4 1035 $iOffset = $i;
8bd0068d 1036 $cur_l = 0;
1037 $ret = '';
1038 $iEncStart = false;
c96c32f4 1039 }
8bd0068d 1040 break;
1041 case ' ':
c96c32f4 1042 if ($iEncStart !== false) {
098ea084 1043 $cur_l++;
1044 if ($cur_l > $max_l) {
08b7f7cc 1045 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
c96c32f4 1046 $aRet[] = "=?$default_charset?Q?$ret?=";
c96c32f4 1047 $iOffset = $i;
1048 $cur_l = 0;
098ea084 1049 $ret = '';
8bd0068d 1050 $iEncStart = false;
08b7f7cc 1051 } else {
8bd0068d 1052 $ret .= '_';
c96c32f4 1053 }
3d8371be 1054 }
8bd0068d 1055 break;
1056 default:
1710ad65 1057 $k = ord($string[$i]);
8bd0068d 1058 if ($k > 126) {
1059 if ($iEncStart === false) {
1060 // do not start encoding in the middle of a string, also take the rest of the word.
1061 $sLeadString = substr($string,0,$i);
1062 $aLeadString = explode(' ',$sLeadString);
1063 $sToBeEncoded = array_pop($aLeadString);
1064 $iEncStart = $i - strlen($sToBeEncoded);
1065 $ret .= $sToBeEncoded;
1066 $cur_l += strlen($sToBeEncoded);
1067 }
1068 $cur_l += 3;
1069 /* first we add the encoded string that reached it's max size */
1070 if ($cur_l > ($max_l-2)) {
1071 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1072 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
1073 $cur_l = 3;
1074 $ret = '';
1075 $iOffset = $i;
1076 $iEncStart = $i;
1077 }
1078 $enc_init = true;
1079 $ret .= sprintf("=%02X", $k);
1080 } else {
1081 if ($iEncStart !== false) {
1082 $cur_l++;
1083 if ($cur_l > $max_l) {
1084 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1085 $aRet[] = "=?$default_charset?Q?$ret?=";
1086 $iEncStart = false;
1087 $iOffset = $i;
1088 $cur_l = 0;
1089 $ret = '';
1090 } else {
1710ad65 1091 $ret .= $string[$i];
8bd0068d 1092 }
1093 }
1094 }
1095 break;
f7b3ba37 1096 }
451f74a2 1097 }
793cc001 1098
c96c32f4 1099 if ($enc_init) {
1100 if ($iEncStart !== false) {
1101 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1102 $aRet[] = "=?$default_charset?Q?$ret?=";
1103 } else {
1104 $aRet[] = substr($string,$iOffset);
1105 }
1106 $string = implode('',$aRet);
451f74a2 1107 }
3d8371be 1108 return $string;
451f74a2 1109}
b74ba498 1110
f270a6eb 1111/**
1112 * Encodes string according to rfc2047 B encoding header formating rules
1113 *
758a7889 1114 * It is recommended way to encode headers with character sets that store
f270a6eb 1115 * symbols in more than one byte.
1116 *
1117 * Function requires mbstring support. If required mbstring functions are missing,
1118 * function returns false and sets E_USER_WARNING level error message.
1119 *
758a7889 1120 * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
1121 * that mbstring functions will generate E_WARNING errors, if unsupported
f270a6eb 1122 * character set is used. mb_encode_mimeheader function provided by php
1123 * mbstring extension is not used in order to get better control of header
1124 * encoding.
1125 *
758a7889 1126 * Used php code functions - function_exists(), trigger_error(), strlen()
1127 * (is used with charset names and base64 strings). Used php mbstring
f270a6eb 1128 * functions - mb_strlen and mb_substr.
1129 *
758a7889 1130 * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
f270a6eb 1131 * encoding), rfc 2822 (header folding)
1132 *
1133 * @param string $string header string that must be encoded
758a7889 1134 * @param string $charset character set. Must be supported by mbstring extension.
f270a6eb 1135 * Use sq_mb_list_encodings() to detect supported charsets.
1136 * @return string string encoded according to rfc2047 B encoding formating rules
1137 * @since 1.5.1
1138 * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1139 * @todo Do we want to control max length of header?
1140 * @todo Do we want to control EOL (end-of-line) marker?
1141 * @todo Do we want to translate error message?
1142 */
1143function encodeHeaderBase64($string,$charset) {
1144 /**
1145 * Check mbstring function requirements.
1146 */
1147 if (! function_exists('mb_strlen') ||
1148 ! function_exists('mb_substr')) {
1149 // set E_USER_WARNING
1150 trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
1151 // return false
1152 return false;
1153 }
1154
1155 // initial return array
1156 $aRet = array();
1157
1158 /**
1159 * header length = 75 symbols max (same as in encodeHeader)
1160 * remove $charset length
1161 * remove =? ? ?= (5 chars)
1162 * remove 2 more chars (\r\n ?)
1163 */
1164 $iMaxLength = 75 - strlen($charset) - 7;
1165
1166 // set first character position
1167 $iStartCharNum = 0;
1168
1169 // loop through all characters. count characters and not bytes.
1170 for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1171 // encode string from starting character to current character.
1172 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1173
1174 // Check encoded string length
1175 if(strlen($encoded_string)>$iMaxLength) {
1176 // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1177 $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1178
1179 // set new starting character
1180 $iStartCharNum = $iCharNum-1;
1181
1182 // encode last char (in case it is last character in string)
1183 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1184 } // if string is shorter than max length - add next character
1185 }
1186
1187 // add last encoded string to array
1188 $aRet[] = $encoded_string;
1189
1190 // set initial return string
1191 $sRet = '';
1192
1193 // loop through encoded strings
1194 foreach($aRet as $string) {
1195 // TODO: Do we want to control EOL (end-of-line) marker
1196 if ($sRet!='') $sRet.= " ";
1197
1198 // add header tags and encoded string to return string
1199 $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1200 }
1201
1202 return $sRet;
1203}
1204
691a2d25 1205/* This function trys to locate the entity_id of a specific mime element */
3d8371be 1206function find_ent_id($id, $message) {
a171b359 1207 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1208 if ($message->entities[$i]->header->type0 == 'multipart') {
3d8371be 1209 $ret = find_ent_id($id, $message->entities[$i]);
451f74a2 1210 } else {
3d8371be 1211 if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
d8cffbab 1212// if (sq_check_save_extension($message->entities[$i])) {
8bd0068d 1213 return $message->entities[$i]->entity_id;
da2415c1 1214// }
c8f5f606 1215 } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1216 /**
1217 * This is part of a fix for Outlook Express 6.x generating
1218 * cid URLs without creating content-id headers
1219 * @@JA - 20050207
1220 */
1221 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1222 return $message->entities[$i]->entity_id;
1223 }
3d8371be 1224 }
a3daaaf3 1225 }
451f74a2 1226 }
3d8371be 1227 return $ret;
451f74a2 1228}
a3daaaf3 1229
e5e9381a 1230function sq_check_save_extension($message) {
1231 $filename = $message->getFilename();
1232 $ext = substr($filename, strrpos($filename,'.')+1);
1233 $save_extensions = array('jpg','jpeg','gif','png','bmp');
3d8371be 1234 return in_array($ext, $save_extensions);
e5e9381a 1235}
1236
1237
691a2d25 1238/**
8bd0068d 1239 ** HTMLFILTER ROUTINES
1240 */
451f74a2 1241
691a2d25 1242/**
0493ed11 1243 * This function checks attribute values for entity-encoded values
1244 * and returns them translated into 8-bit strings so we can run
1245 * checks on them.
8bd0068d 1246 *
0493ed11 1247 * @param $attvalue A string to run entity check against.
1248 * @return Nothing, modifies a reference value.
8bd0068d 1249 */
0493ed11 1250function sq_defang(&$attvalue){
1251 $me = 'sq_defang';
2dd879b8 1252 /**
0493ed11 1253 * Skip this if there aren't ampersands or backslashes.
8bd0068d 1254 */
0493ed11 1255 if (strpos($attvalue, '&') === false
1256 && strpos($attvalue, '\\') === false){
1257 return;
2dd879b8 1258 }
0493ed11 1259 $m = false;
2b665f28 1260 // before deent, translate the dangerous unicode characters and ... to safe values
1261 // otherwise the regular expressions do not match.
1262
1263
1264
0493ed11 1265 do {
1266 $m = false;
1267 $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1268 $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1269 $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1270 } while ($m == true);
1271 $attvalue = stripslashes($attvalue);
2dd879b8 1272}
1273
1274/**
8bd0068d 1275 * Kill any tabs, newlines, or carriage returns. Our friends the
1276 * makers of the browser with 95% market value decided that it'd
1277 * be funny to make "java[tab]script" be just as good as "javascript".
1278 *
1279 * @param attvalue The attribute value before extraneous spaces removed.
0493ed11 1280 * @return attvalue Nothing, modifies a reference value.
8bd0068d 1281 */
0493ed11 1282function sq_unspace(&$attvalue){
1283 $me = 'sq_unspace';
1284 if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1285 $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1286 Array('', '', '', '', ''), $attvalue);
2dd879b8 1287 }
2dd879b8 1288}
1289
1290/**
88de4926 1291 * Translate all dangerous Unicode or Shift_JIS characters which are accepted by
2b665f28 1292 * IE as regular characters.
1293 *
1294 * @param attvalue The attribute value before dangerous characters are translated.
1295 * @return attvalue Nothing, modifies a reference value.
1296 * @author Marc Groot Koerkamp.
1297 */
1298function sq_fixIE_idiocy(&$attvalue) {
1299 // remove NUL
1300 $attvalue = str_replace("\0", "", $attvalue);
1301 // remove comments
1302 $attvalue = preg_replace("/(\/\*.*?\*\/)/","",$attvalue);
1303
88de4926 1304 // IE has the evil habit of accepting every possible value for the attribute expression.
1305 // The table below contains characters which are parsed by IE if they are used in the "expression"
2b665f28 1306 // attribute value.
1307 $aDangerousCharsReplacementTable = array(
1308 array('&#x029F;', '&#0671;' ,/* L UNICODE IPA Extension */
1309 '&#x0280;', '&#0640;' ,/* R UNICODE IPA Extension */
1310 '&#x0274;', '&#0628;' ,/* N UNICODE IPA Extension */
567dc524 1311 '&#xFF25;', '&#65317;' ,/* Unicode FULLWIDTH LATIN CAPITAL LETTER E */
1312 '&#xFF45;', '&#65349;' ,/* Unicode FULLWIDTH LATIN SMALL LETTER E */
2b665f28 1313 '&#xFF38;', '&#65336;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER X */
1314 '&#xFF58;', '&#65368;',/* Unicode FULLWIDTH LATIN SMALL LETTER X */
1315 '&#xFF30;', '&#65328;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER P */
1316 '&#xFF50;', '&#65360;',/* Unicode FULLWIDTH LATIN SMALL LETTER P */
1317 '&#xFF32;', '&#65330;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER R */
1318 '&#xFF52;', '&#65362;',/* Unicode FULLWIDTH LATIN SMALL LETTER R */
1319 '&#xFF33;', '&#65331;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER S */
1320 '&#xFF53;', '&#65363;',/* Unicode FULLWIDTH LATIN SMALL LETTER S */
1321 '&#xFF29;', '&#65321;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER I */
1322 '&#xFF49;', '&#65353;',/* Unicode FULLWIDTH LATIN SMALL LETTER I */
1323 '&#xFF2F;', '&#65327;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER O */
1324 '&#xFF4F;', '&#65359;',/* Unicode FULLWIDTH LATIN SMALL LETTER O */
1325 '&#xFF2E;', '&#65326;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER N */
1326 '&#xFF4E;', '&#65358;',/* Unicode FULLWIDTH LATIN SMALL LETTER N */
1327 '&#xFF2C;', '&#65324;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER L */
1328 '&#xFF4C;', '&#65356;',/* Unicode FULLWIDTH LATIN SMALL LETTER L */
1329 '&#xFF35;', '&#65333;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER U */
1330 '&#xFF55;', '&#65365;',/* Unicode FULLWIDTH LATIN SMALL LETTER U */
1331 '&#x207F;', '&#8319;' ,/* Unicode SUPERSCRIPT LATIN SMALL LETTER N */
567dc524 1332 "\xEF\xBC\xA5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER E */ // in unicode this is some Chinese char range
1333 "\xEF\xBD\x85", /* Shift JIS FULLWIDTH LATIN SMALL LETTER E */
1334 "\xEF\xBC\xB8", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER X */
1335 "\xEF\xBD\x98", /* Shift JIS FULLWIDTH LATIN SMALL LETTER X */
1336 "\xEF\xBC\xB0", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER P */
1337 "\xEF\xBD\x90", /* Shift JIS FULLWIDTH LATIN SMALL LETTER P */
1338 "\xEF\xBC\xB2", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER R */
1339 "\xEF\xBD\x92", /* Shift JIS FULLWIDTH LATIN SMALL LETTER R */
1340 "\xEF\xBC\xB3", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER S */
1341 "\xEF\xBD\x93", /* Shift JIS FULLWIDTH LATIN SMALL LETTER S */
1342 "\xEF\xBC\xA9", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER I */
1343 "\xEF\xBD\x89", /* Shift JIS FULLWIDTH LATIN SMALL LETTER I */
1344 "\xEF\xBC\xAF", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER O */
1345 "\xEF\xBD\x8F", /* Shift JIS FULLWIDTH LATIN SMALL LETTER O */
1346 "\xEF\xBC\xAE", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER N */
1347 "\xEF\xBD\x8E", /* Shift JIS FULLWIDTH LATIN SMALL LETTER N */
1348 "\xEF\xBC\xAC", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER L */
1349 "\xEF\xBD\x8C", /* Shift JIS FULLWIDTH LATIN SMALL LETTER L */
1350 "\xEF\xBC\xB5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER U */
1351 "\xEF\xBD\x95", /* Shift JIS FULLWIDTH LATIN SMALL LETTER U */
1352 "\xE2\x81\xBF", /* Shift JIS FULLWIDTH SUPERSCRIPT N */
1353 "\xCA\x9F", /* L UNICODE IPA Extension */
1354 "\xCA\x80", /* R UNICODE IPA Extension */
1355 "\xC9\xB4"), /* N UNICODE IPA Extension */
2b665f28 1356 array('l', 'l', 'r','r','n','n',
567dc524 1357 'E','E','e','e','X','X','x','x','P','P','p','p','R','R','r','r','S','S','s','s','I','I',
1358 'i','i','O','O','o','o','N','N','n','n','L','L','l','l','U','U','u','u','n','n',
1359 '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 1360 $attvalue = str_replace($aDangerousCharsReplacementTable[0],$aDangerousCharsReplacementTable[1],$attvalue);
1361
88de4926 1362 // Escapes are useful for special characters like "{}[]()'&. In other cases they are
1363 // used for XSS.
2b665f28 1364 $attvalue = preg_replace("/(\\\\)([a-zA-Z]{1})/",'$2',$attvalue);
1365}
1366
1367/**
8bd0068d 1368 * This function returns the final tag out of the tag name, an array
1369 * of attributes, and the type of the tag. This function is called by
1370 * sq_sanitize internally.
1371 *
1372 * @param $tagname the name of the tag.
1373 * @param $attary the array of attributes and their values
1374 * @param $tagtype The type of the tag (see in comments).
1375 * @return a string with the final tag representation.
1376 */
691a2d25 1377function sq_tagprint($tagname, $attary, $tagtype){
b583c3e8 1378 $me = 'sq_tagprint';
3d8371be 1379
691a2d25 1380 if ($tagtype == 2){
1381 $fulltag = '</' . $tagname . '>';
1382 } else {
1383 $fulltag = '<' . $tagname;
1384 if (is_array($attary) && sizeof($attary)){
1385 $atts = Array();
c71c845c 1386 foreach ($attary as $attname => $attvalue){
691a2d25 1387 array_push($atts, "$attname=$attvalue");
1388 }
1389 $fulltag .= ' ' . join(" ", $atts);
1390 }
1391 if ($tagtype == 3){
b583c3e8 1392 $fulltag .= ' /';
691a2d25 1393 }
b583c3e8 1394 $fulltag .= '>';
451f74a2 1395 }
691a2d25 1396 return $fulltag;
451f74a2 1397}
a3daaaf3 1398
691a2d25 1399/**
8bd0068d 1400 * A small helper function to use with array_walk. Modifies a by-ref
1401 * value and makes it lowercase.
1402 *
1403 * @param $val a value passed by-ref.
1404 * @return void since it modifies a by-ref value.
1405 */
691a2d25 1406function sq_casenormalize(&$val){
1407 $val = strtolower($val);
1408}
451f74a2 1409
691a2d25 1410/**
8bd0068d 1411 * This function skips any whitespace from the current position within
1412 * a string and to the next non-whitespace value.
1413 *
1414 * @param $body the string
1415 * @param $offset the offset within the string where we should start
1416 * looking for the next non-whitespace character.
1417 * @return the location within the $body where the next
1418 * non-whitespace char is located.
1419 */
691a2d25 1420function sq_skipspace($body, $offset){
b583c3e8 1421 $me = 'sq_skipspace';
3d8371be 1422 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
aab65dbd 1423 if (!empty($matches[1])){
1424 $offset += strlen($matches[1]);
451f74a2 1425 }
691a2d25 1426 return $offset;
451f74a2 1427}
a3daaaf3 1428
691a2d25 1429/**
8bd0068d 1430 * This function looks for the next character within a string. It's
1431 * really just a glorified "strpos", except it catches if failures
1432 * nicely.
1433 *
1434 * @param $body The string to look for needle in.
1435 * @param $offset Start looking from this position.
1436 * @param $needle The character/string to look for.
1437 * @return location of the next occurance of the needle, or
1438 * strlen($body) if needle wasn't found.
1439 */
691a2d25 1440function sq_findnxstr($body, $offset, $needle){
3d8371be 1441 $me = 'sq_findnxstr';
691a2d25 1442 $pos = strpos($body, $needle, $offset);
1443 if ($pos === FALSE){
1444 $pos = strlen($body);
451f74a2 1445 }
691a2d25 1446 return $pos;
451f74a2 1447}
a3daaaf3 1448
691a2d25 1449/**
8bd0068d 1450 * This function takes a PCRE-style regexp and tries to match it
1451 * within the string.
1452 *
1453 * @param $body The string to look for needle in.
1454 * @param $offset Start looking from here.
1455 * @param $reg A PCRE-style regex to match.
1456 * @return Returns a false if no matches found, or an array
1457 * with the following members:
1458 * - integer with the location of the match within $body
1459 * - string with whatever content between offset and the match
1460 * - string with whatever it is we matched
1461 */
691a2d25 1462function sq_findnxreg($body, $offset, $reg){
b583c3e8 1463 $me = 'sq_findnxreg';
691a2d25 1464 $matches = Array();
1465 $retarr = Array();
7d06541f 1466 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1710ad65 1467 if (!isset($matches[0]) || !$matches[0]){
691a2d25 1468 $retarr = false;
1469 } else {
1710ad65 1470 $retarr[0] = $offset + strlen($matches[1]);
1471 $retarr[1] = $matches[1];
1472 $retarr[2] = $matches[2];
691a2d25 1473 }
1474 return $retarr;
1475}
a3daaaf3 1476
691a2d25 1477/**
8bd0068d 1478 * This function looks for the next tag.
1479 *
1480 * @param $body String where to look for the next tag.
1481 * @param $offset Start looking from here.
1482 * @return false if no more tags exist in the body, or
1483 * an array with the following members:
1484 * - string with the name of the tag
1485 * - array with attributes and their values
1486 * - integer with tag type (1, 2, or 3)
1487 * - integer where the tag starts (starting "<")
1488 * - integer where the tag ends (ending ">")
1489 * first three members will be false, if the tag is invalid.
1490 */
691a2d25 1491function sq_getnxtag($body, $offset){
b583c3e8 1492 $me = 'sq_getnxtag';
691a2d25 1493 if ($offset > strlen($body)){
1494 return false;
1495 }
1496 $lt = sq_findnxstr($body, $offset, "<");
1497 if ($lt == strlen($body)){
1498 return false;
1499 }
1500 /**
8bd0068d 1501 * We are here:
1502 * blah blah <tag attribute="value">
1503 * \---------^
1504 */
691a2d25 1505 $pos = sq_skipspace($body, $lt+1);
1506 if ($pos >= strlen($body)){
1507 return Array(false, false, false, $lt, strlen($body));
1508 }
1509 /**
8bd0068d 1510 * There are 3 kinds of tags:
1511 * 1. Opening tag, e.g.:
1512 * <a href="blah">
1513 * 2. Closing tag, e.g.:
1514 * </a>
1515 * 3. XHTML-style content-less tag, e.g.:
1516 * <img src="blah" />
1517 */
691a2d25 1518 $tagtype = false;
1519 switch (substr($body, $pos, 1)){
3d8371be 1520 case '/':
1521 $tagtype = 2;
1522 $pos++;
1523 break;
1524 case '!':
1525 /**
8bd0068d 1526 * A comment or an SGML declaration.
1527 */
3d8371be 1528 if (substr($body, $pos+1, 2) == "--"){
1529 $gt = strpos($body, "-->", $pos);
1530 if ($gt === false){
1531 $gt = strlen($body);
1532 } else {
1533 $gt += 2;
1534 }
1535 return Array(false, false, false, $lt, $gt);
bb8d0799 1536 } else {
3d8371be 1537 $gt = sq_findnxstr($body, $pos, ">");
1538 return Array(false, false, false, $lt, $gt);
1539 }
1540 break;
1541 default:
1542 /**
8bd0068d 1543 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1544 * later.
1545 */
3d8371be 1546 $tagtype = 1;
1547 break;
691a2d25 1548 }
a3daaaf3 1549
0493ed11 1550 $tag_start = $pos;
691a2d25 1551 $tagname = '';
1552 /**
8bd0068d 1553 * Look for next [\W-_], which will indicate the end of the tag name.
1554 */
691a2d25 1555 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1556 if ($regary == false){
1557 return Array(false, false, false, $lt, strlen($body));
1558 }
1559 list($pos, $tagname, $match) = $regary;
1560 $tagname = strtolower($tagname);
1561
1562 /**
8bd0068d 1563 * $match can be either of these:
1564 * '>' indicating the end of the tag entirely.
1565 * '\s' indicating the end of the tag name.
1566 * '/' indicating that this is type-3 xhtml tag.
1567 *
1568 * Whatever else we find there indicates an invalid tag.
1569 */
691a2d25 1570 switch ($match){
3d8371be 1571 case '/':
691a2d25 1572 /**
8bd0068d 1573 * This is an xhtml-style tag with a closing / at the
1574 * end, like so: <img src="blah" />. Check if it's followed
1575 * by the closing bracket. If not, then this tag is invalid
1576 */
3d8371be 1577 if (substr($body, $pos, 2) == "/>"){
1578 $pos++;
1579 $tagtype = 3;
1580 } else {
1581 $gt = sq_findnxstr($body, $pos, ">");
1582 $retary = Array(false, false, false, $lt, $gt);
1583 return $retary;
1584 }
1585 case '>':
1586 return Array($tagname, false, $tagtype, $lt, $pos);
1587 break;
1588 default:
1589 /**
8bd0068d 1590 * Check if it's whitespace
1591 */
3d8371be 1592 if (!preg_match('/\s/', $match)){
1593 /**
8bd0068d 1594 * This is an invalid tag! Look for the next closing ">".
1595 */
7d06541f 1596 $gt = sq_findnxstr($body, $lt, ">");
3d8371be 1597 return Array(false, false, false, $lt, $gt);
1598 }
1599 break;
691a2d25 1600 }
3d8371be 1601
691a2d25 1602 /**
8bd0068d 1603 * At this point we're here:
1604 * <tagname attribute='blah'>
1605 * \-------^
1606 *
1607 * At this point we loop in order to find all attributes.
1608 */
691a2d25 1609 $attname = '';
0493ed11 1610 $atttype = false;
691a2d25 1611 $attary = Array();
1612
1613 while ($pos <= strlen($body)){
1614 $pos = sq_skipspace($body, $pos);
1615 if ($pos == strlen($body)){
1616 /**
8bd0068d 1617 * Non-closed tag.
1618 */
691a2d25 1619 return Array(false, false, false, $lt, $pos);
1620 }
1621 /**
8bd0068d 1622 * See if we arrived at a ">" or "/>", which means that we reached
1623 * the end of the tag.
1624 */
691a2d25 1625 $matches = Array();
164800ad 1626 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
c828931c 1627 /**
8bd0068d 1628 * Yep. So we did.
1629 */
1710ad65 1630 $pos += strlen($matches[1]);
1631 if ($matches[2] == "/>"){
c828931c 1632 $tagtype = 3;
1633 $pos++;
1634 }
1635 return Array($tagname, $attary, $tagtype, $lt, $pos);
1636 }
a3daaaf3 1637
cca46357 1638 /**
8bd0068d 1639 * There are several types of attributes, with optional
1640 * [:space:] between members.
1641 * Type 1:
1642 * attrname[:space:]=[:space:]'CDATA'
1643 * Type 2:
1644 * attrname[:space:]=[:space:]"CDATA"
1645 * Type 3:
1646 * attr[:space:]=[:space:]CDATA
1647 * Type 4:
1648 * attrname
1649 *
1650 * We leave types 1 and 2 the same, type 3 we check for
1651 * '"' and convert to "&quot" if needed, then wrap in
1652 * double quotes. Type 4 we convert into:
1653 * attrname="yes".
1654 */
3f7c623f 1655 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
691a2d25 1656 if ($regary == false){
1657 /**
8bd0068d 1658 * Looks like body ended before the end of tag.
1659 */
691a2d25 1660 return Array(false, false, false, $lt, strlen($body));
cca46357 1661 }
691a2d25 1662 list($pos, $attname, $match) = $regary;
1663 $attname = strtolower($attname);
1664 /**
8bd0068d 1665 * We arrived at the end of attribute name. Several things possible
1666 * here:
1667 * '>' means the end of the tag and this is attribute type 4
1668 * '/' if followed by '>' means the same thing as above
1669 * '\s' means a lot of things -- look what it's followed by.
1670 * anything else means the attribute is invalid.
1671 */
691a2d25 1672 switch($match){
3d8371be 1673 case '/':
691a2d25 1674 /**
8bd0068d 1675 * This is an xhtml-style tag with a closing / at the
1676 * end, like so: <img src="blah" />. Check if it's followed
1677 * by the closing bracket. If not, then this tag is invalid
1678 */
3d8371be 1679 if (substr($body, $pos, 2) == "/>"){
691a2d25 1680 $pos++;
3d8371be 1681 $tagtype = 3;
691a2d25 1682 } else {
3d8371be 1683 $gt = sq_findnxstr($body, $pos, ">");
1684 $retary = Array(false, false, false, $lt, $gt);
1685 return $retary;
1686 }
1687 case '>':
1710ad65 1688 $attary[$attname] = '"yes"';
3d8371be 1689 return Array($tagname, $attary, $tagtype, $lt, $pos);
1690 break;
1691 default:
1692 /**
8bd0068d 1693 * Skip whitespace and see what we arrive at.
1694 */
3d8371be 1695 $pos = sq_skipspace($body, $pos);
1696 $char = substr($body, $pos, 1);
1697 /**
8bd0068d 1698 * Two things are valid here:
1699 * '=' means this is attribute type 1 2 or 3.
1700 * \w means this was attribute type 4.
1701 * anything else we ignore and re-loop. End of tag and
1702 * invalid stuff will be caught by our checks at the beginning
1703 * of the loop.
1704 */
3d8371be 1705 if ($char == "="){
1706 $pos++;
1707 $pos = sq_skipspace($body, $pos);
691a2d25 1708 /**
8bd0068d 1709 * Here are 3 possibilities:
1710 * "'" attribute type 1
1711 * '"' attribute type 2
1712 * everything else is the content of tag type 3
1713 */
3d8371be 1714 $quot = substr($body, $pos, 1);
1715 if ($quot == "'"){
1716 $regary = sq_findnxreg($body, $pos+1, "\'");
1717 if ($regary == false){
1718 return Array(false, false, false, $lt, strlen($body));
1719 }
1720 list($pos, $attval, $match) = $regary;
1721 $pos++;
1710ad65 1722 $attary[$attname] = "'" . $attval . "'";
3d8371be 1723 } else if ($quot == '"'){
1724 $regary = sq_findnxreg($body, $pos+1, '\"');
1725 if ($regary == false){
1726 return Array(false, false, false, $lt, strlen($body));
1727 }
1728 list($pos, $attval, $match) = $regary;
1729 $pos++;
1710ad65 1730 $attary[$attname] = '"' . $attval . '"';
3d8371be 1731 } else {
1732 /**
8bd0068d 1733 * These are hateful. Look for \s, or >.
1734 */
3d8371be 1735 $regary = sq_findnxreg($body, $pos, "[\s>]");
1736 if ($regary == false){
1737 return Array(false, false, false, $lt, strlen($body));
1738 }
1739 list($pos, $attval, $match) = $regary;
1740 /**
8bd0068d 1741 * If it's ">" it will be caught at the top.
1742 */
3d8371be 1743 $attval = preg_replace("/\"/s", "&quot;", $attval);
1710ad65 1744 $attary[$attname] = '"' . $attval . '"';
7e235a1a 1745 }
3d8371be 1746 } else if (preg_match("|[\w/>]|", $char)) {
691a2d25 1747 /**
8bd0068d 1748 * That was attribute type 4.
1749 */
1710ad65 1750 $attary[$attname] = '"yes"';
3d8371be 1751 } else {
1752 /**
8bd0068d 1753 * An illegal character. Find next '>' and return.
1754 */
3d8371be 1755 $gt = sq_findnxstr($body, $pos, ">");
1756 return Array(false, false, false, $lt, $gt);
451f74a2 1757 }
3d8371be 1758 break;
691a2d25 1759 }
1760 }
1761 /**
8bd0068d 1762 * The fact that we got here indicates that the tag end was never
1763 * found. Return invalid tag indication so it gets stripped.
1764 */
691a2d25 1765 return Array(false, false, false, $lt, strlen($body));
1766}
1767
1768/**
0493ed11 1769 * Translates entities into literal values so they can be checked.
8bd0068d 1770 *
0493ed11 1771 * @param $attvalue the by-ref value to check.
1772 * @param $regex the regular expression to check against.
1773 * @param $hex whether the entites are hexadecimal.
1774 * @return True or False depending on whether there were matches.
8bd0068d 1775 */
0493ed11 1776function sq_deent(&$attvalue, $regex, $hex=false){
b583c3e8 1777 $me = 'sq_deent';
0493ed11 1778 $ret_match = false;
2b665f28 1779 // remove comments
1780 //$attvalue = preg_replace("/(\/\*.*\*\/)/","",$attvalue);
0493ed11 1781 preg_match_all($regex, $attvalue, $matches);
1782 if (is_array($matches) && sizeof($matches[0]) > 0){
1783 $repl = Array();
1784 for ($i = 0; $i < sizeof($matches[0]); $i++){
1785 $numval = $matches[1][$i];
1786 if ($hex){
1787 $numval = hexdec($numval);
a3daaaf3 1788 }
1710ad65 1789 $repl[$matches[0][$i]] = chr($numval);
691a2d25 1790 }
0493ed11 1791 $attvalue = strtr($attvalue, $repl);
1792 return true;
1793 } else {
1794 return false;
691a2d25 1795 }
691a2d25 1796}
1797
1798/**
8bd0068d 1799 * This function runs various checks against the attributes.
1800 *
1801 * @param $tagname String with the name of the tag.
1802 * @param $attary Array with all tag attributes.
1803 * @param $rm_attnames See description for sq_sanitize
1804 * @param $bad_attvals See description for sq_sanitize
1805 * @param $add_attr_to_tag See description for sq_sanitize
1806 * @param $message message object
1807 * @param $id message id
1808 * @return Array with modified attributes.
1809 */
da2415c1 1810function sq_fixatts($tagname,
1811 $attary,
691a2d25 1812 $rm_attnames,
1813 $bad_attvals,
1814 $add_attr_to_tag,
1815 $message,
b3af12ef 1816 $id,
3d8371be 1817 $mailbox
691a2d25 1818 ){
b583c3e8 1819 $me = 'sq_fixatts';
c71c845c 1820 foreach ($attary as $attname => $attvalue){
691a2d25 1821 /**
8bd0068d 1822 * See if this attribute should be removed.
1823 */
691a2d25 1824 foreach ($rm_attnames as $matchtag=>$matchattrs){
1825 if (preg_match($matchtag, $tagname)){
1826 foreach ($matchattrs as $matchattr){
1827 if (preg_match($matchattr, $attname)){
1710ad65 1828 unset($attary[$attname]);
691a2d25 1829 continue;
1830 }
451f74a2 1831 }
451f74a2 1832 }
691a2d25 1833 }
1834 /**
2b665f28 1835 * Workaround for IE quirks
1836 */
1837 sq_fixIE_idiocy($attvalue);
1838
1839 /**
8bd0068d 1840 * Remove any backslashes, entities, and extraneous whitespace.
1841 */
2b665f28 1842
1843 $oldattvalue = $attvalue;
0493ed11 1844 sq_defang($attvalue);
2b665f28 1845 if ($attname == 'style' && $attvalue !== $oldattvalue) {
1846 // entities are used in the attribute value. In 99% of the cases it's there as XSS
1847 // i.e.<div style="{ left:exp&#x0280;essio&#x0274;( alert('XSS') ) }">
1848 $attvalue = "idiocy";
1710ad65 1849 $attary[$attname] = $attvalue;
2b665f28 1850 }
0493ed11 1851 sq_unspace($attvalue);
af861a34 1852
1853 /**
8bd0068d 1854 * Now let's run checks on the attvalues.
1855 * I don't expect anyone to comprehend this. If you do,
1856 * get in touch with me so I can drive to where you live and
1857 * shake your hand personally. :)
1858 */
691a2d25 1859 foreach ($bad_attvals as $matchtag=>$matchattrs){
1860 if (preg_match($matchtag, $tagname)){
1861 foreach ($matchattrs as $matchattr=>$valary){
1862 if (preg_match($matchattr, $attname)){
1863 /**
8bd0068d 1864 * There are two arrays in valary.
1865 * First is matches.
1866 * Second one is replacements
1867 */
691a2d25 1868 list($valmatch, $valrepl) = $valary;
da2415c1 1869 $newvalue =
691a2d25 1870 preg_replace($valmatch, $valrepl, $attvalue);
1871 if ($newvalue != $attvalue){
1710ad65 1872 $attary[$attname] = $newvalue;
567dc524 1873 $attvalue = $newvalue;
691a2d25 1874 }
1875 }
1876 }
451f74a2 1877 }
a3daaaf3 1878 }
567dc524 1879 if ($attname == 'style') {
1880 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1881 // 8bit and control characters in style attribute values can be used for XSS, remove them
1710ad65 1882 $attary[$attname] = '"disallowed character"';
567dc524 1883 }
1884 preg_match_all("/url\s*\((.+)\)/si",$attvalue,$aMatch);
1885 if (count($aMatch)) {
1886 foreach($aMatch[1] as $sMatch) {
1887 // url value
1888 $urlvalue = $sMatch;
1889 sq_fix_url($attname, $urlvalue, $message, $id, $mailbox,"'");
1710ad65 1890 $attary[$attname] = str_replace($sMatch,$urlvalue,$attvalue);
567dc524 1891 }
1892 }
691a2d25 1893 }
ff940ebc 1894 /**
567dc524 1895 * Use white list based filtering on attributes which can contain url's
ff940ebc 1896 */
caf0ab1d 1897 else if ($attname == 'href' || $attname == 'xlink:href' || $attname == 'src'
1898 || $attname == 'poster' || $attname == 'formaction'
1899 || $attname == 'background' || $attname == 'action') {
567dc524 1900 sq_fix_url($attname, $attvalue, $message, $id, $mailbox);
1710ad65 1901 $attary[$attname] = $attvalue;
ff940ebc 1902 }
a3daaaf3 1903 }
691a2d25 1904 /**
8bd0068d 1905 * See if we need to append any attributes to this tag.
1906 */
691a2d25 1907 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1908 if (preg_match($matchtag, $tagname)){
1909 $attary = array_merge($attary, $addattary);
1910 }
1911 }
1912 return $attary;
451f74a2 1913}
a3daaaf3 1914
691a2d25 1915/**
567dc524 1916 * This function filters url's
1917 *
1918 * @param $attvalue String with attribute value to filter
1919 * @param $message message object
1920 * @param $id message id
1921 * @param $mailbox mailbox
1922 * @param $sQuote quoting characters around url's
1923 */
1924function sq_fix_url($attname, &$attvalue, $message, $id, $mailbox,$sQuote = '"') {
1925 $attvalue = trim($attvalue);
1926 if ($attvalue && ($attvalue[0] =='"'|| $attvalue[0] == "'")) {
1927 // remove the double quotes
1928 $sQuote = $attvalue[0];
1929 $attvalue = trim(substr($attvalue,1,-1));
1930 }
1931
d31f73f1 1932 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
1933 // images off by default.
afbf184c 1934 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET, FALSE);
d31f73f1 1935
6ab8321f 1936 global $use_transparent_security_image;
1937 if ($use_transparent_security_image) $secremoveimg = '../images/spacer.png';
1938 else $secremoveimg = '../images/' . _("sec_remove_eng.png");
567dc524 1939
1940 /**
1941 * Replace empty src tags with the blank image. src is only used
1942 * for frames, images, and image inputs. Doing a replace should
1943 * not affect them working as should be, however it will stop
1944 * IE from being kicked off when src for img tags are not set
1945 */
1946 if ($attvalue == '') {
1947 $attvalue = '"' . SM_PATH . 'images/blank.png"';
1948 } else {
1949 // first, disallow 8 bit characters and control characters
1950 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1951 switch ($attname) {
1952 case 'href':
1953 $attvalue = $sQuote . 'http://invalid-stuff-detected.example.com' . $sQuote;
1954 break;
1955 default:
1956 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1957 break;
1958 }
1959 } else {
1960 $aUrl = parse_url($attvalue);
1961 if (isset($aUrl['scheme'])) {
1962 switch(strtolower($aUrl['scheme'])) {
d75e755b 1963 case 'mailto':
567dc524 1964 case 'http':
1965 case 'https':
1966 case 'ftp':
1967 if ($attname != 'href') {
1968 if ($view_unsafe_images == false) {
1969 $attvalue = $sQuote . $secremoveimg . $sQuote;
1970 } else {
1971 if (isset($aUrl['path'])) {
8bbbc757 1972
1973 // No one has been able to show that image URIs
1974 // can be exploited, so for now, no restrictions
1975 // are made at all. If this proves to be a problem,
1976 // the commented-out code below can be of help.
1977 // (One consideration is that I see nothing in this
1978 // function that specifically says that we will
1979 // only ever arrive here when inspecting an image
1980 // tag, although that does seem to be the end
1981 // result - e.g., <script src="..."> where malicious
1982 // image URIs are in fact a problem are already
1983 // filtered out elsewhere.
1984 /* ---------------------------------
567dc524 1985 // validate image extension.
1986 $ext = strtolower(substr($aUrl['path'],strrpos($aUrl['path'],'.')));
1987 if (!in_array($ext,array('.jpeg','.jpg','xjpeg','.gif','.bmp','.jpe','.png','.xbm'))) {
8bbbc757 1988 // If URI is to something other than
1989 // a regular image file, get the contents
1990 // and try to see if it is an image.
1991 // Don't use Fileinfo (finfo_file()) because
1992 // we'd need to make the admin configure the
1993 // location of the magic.mime file (FIXME: add finfo_file() support later?)
1994 //
1995 $mime_type = '';
1996 if (function_exists('mime_content_type')
1997 && ($FILE = @fopen($attvalue, 'rb', FALSE))) {
1998
1999 // fetch file
2000 //
2001 $file_contents = '';
2002 while (!feof($FILE)) {
2003 $file_contents .= fread($FILE, 8192);
2004 }
2005 fclose($FILE);
2006
2007 // store file locally
2008 //
2009 global $attachment_dir, $username;
2010 $hashed_attachment_dir = getHashedDir($username, $attachment_dir);
2011 $localfilename = GenerateRandomString(32, '', 7);
2012 $full_localfilename = "$hashed_attachment_dir/$localfilename";
2013 while (file_exists($full_localfilename)) {
2014 $localfilename = GenerateRandomString(32, '', 7);
2015 $full_localfilename = "$hashed_attachment_dir/$localfilename";
2016 }
2017 $FILE = fopen("$hashed_attachment_dir/$localfilename", 'wb');
2018 fwrite($FILE, $file_contents);
2019 fclose($FILE);
2020
2021 // get mime type and remove file
2022 //
2023 $mime_type = mime_content_type("$hashed_attachment_dir/$localfilename");
2024 unlink("$hashed_attachment_dir/$localfilename");
2025 }
2026 // debug: echo "$attvalue FILE TYPE IS $mime_type<HR>";
2027 if (substr(strtolower($mime_type), 0, 5) != 'image') {
2028 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
2029 }
567dc524 2030 }
8bbbc757 2031 --------------------------------- */
567dc524 2032 } else {
2033 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
2034 }
2035 }
cf088f55 2036 } else {
2037 $attvalue = $sQuote . $attvalue . $sQuote;
567dc524 2038 }
2039 break;
2040 case 'outbind':
2041 /**
2042 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
2043 * One day MS might actually make it match something useful, for now, falling
2044 * back to using cid2http, so we can grab the blank.png.
2045 */
cf088f55 2046 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
567dc524 2047 break;
2048 case 'cid':
2049 /**
2050 * Turn cid: urls into http-friendly ones.
2051 */
cf088f55 2052 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
567dc524 2053 break;
2054 default:
2055 $attvalue = $sQuote . SM_PATH . 'images/blank.png' . $sQuote;
2056 break;
2057 }
2058 } else {
b1fbb25f 2059 if (!isset($aUrl['path']) || $aUrl['path'] != $secremoveimg) {
567dc524 2060 // parse_url did not lead to satisfying result
2061 $attvalue = $sQuote . SM_PATH . 'images/blank.png' . $sQuote;
2062 }
2063 }
2064 }
2065 }
2066}
2067
2068/**
8bd0068d 2069 * This function edits the style definition to make them friendly and
2070 * usable in SquirrelMail.
2071 *
2072 * @param $message the message object
2073 * @param $id the message id
2074 * @param $content a string with whatever is between <style> and </style>
2075 * @param $mailbox the message mailbox
2076 * @return a string with edited content.
2077 */
e60a299a 2078function sq_fixstyle($body, $pos, $message, $id, $mailbox){
b583c3e8 2079 $me = 'sq_fixstyle';
7e2ff844 2080 // workaround for </style> in between comments
2081 $iCurrentPos = $pos;
2082 $content = '';
2083 $sToken = '';
2084 $bSucces = false;
2085 $bEndTag = false;
2086 for ($i=$pos,$iCount=strlen($body);$i<$iCount;++$i) {
1710ad65 2087 $char = $body[$i];
7e2ff844 2088 switch ($char) {
2089 case '<':
10c0caeb 2090 $sToken = $char;
7e2ff844 2091 break;
2092 case '/':
2093 if ($sToken == '<') {
2094 $sToken .= $char;
2095 $bEndTag = true;
2096 } else {
2097 $content .= $char;
2098 }
2099 break;
2100 case '>':
2101 if ($bEndTag) {
2102 $sToken .= $char;
2103 if (preg_match('/\<\/\s*style\s*\>/i',$sToken,$aMatch)) {
2104 $newpos = $i + 1;
2105 $bSucces = true;
2106 break 2;
2107 } else {
2108 $content .= $sToken;
2109 }
2110 $bEndTag = false;
2111 } else {
2112 $content .= $char;
2113 }
2114 break;
2115 case '!':
2116 if ($sToken == '<') {
2117 // possible comment
1710ad65 2118 if (isset($body[$i+2]) && substr($body,$i,3) == '!--') {
7e2ff844 2119 $i = strpos($body,'-->',$i+3);
2b665f28 2120 if ($i === false) { // no end comment
2121 $i = strlen($body);
2122 }
7e2ff844 2123 $sToken = '';
2124 }
2125 } else {
2126 $content .= $char;
2127 }
2128 break;
2129 default:
2130 if ($bEndTag) {
2131 $sToken .= $char;
2132 } else {
2133 $content .= $char;
2134 }
2135 break;
2136 }
2137 }
2138 if ($bSucces == FALSE){
7d06541f 2139 return array(FALSE, strlen($body));
2140 }
7e2ff844 2141
2b665f28 2142
2143
691a2d25 2144 /**
9ebace19 2145 * First look for general BODY style declaration, which would be
2146 * like so:
2147 * body {background: blah-blah}
2148 * and change it to .bodyclass so we can just assign it to a <div>
2149 */
7732acc6 2150 // $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
d2f3a6fb 2151 // Nah, this is even better - try to preface all CSS selectors with
2152 // our <div> class ID "bodyclass" then correct generic "body" selectors
2153 // TODO: this works pretty good but breaks stuff like this:
2154 // @media print { body { font-size: 10pt; } }
2155 // but there isn't an easy way to make this regex skip @media
2156 // definitions... though lots of the ones in the wild will be
2157 // correctly handled because they tend to end with a parenthesis, like:
2158 // @media screen and (max-width:480px) { ...
2159 $content = preg_replace('/([a-z0-9._-][a-z0-9 >+~|:._-]*\s*(?:,|{.*?}))/si', '.bodyclass $1', $content);
2160 $content = str_replace('.bodyclass body', '.bodyclass', $content);
6ab8321f 2161
2162 global $use_transparent_security_image;
2163 if ($use_transparent_security_image) $secremoveimg = '../images/spacer.png';
2164 else $secremoveimg = '../images/' . _("sec_remove_eng.png");
2165
691a2d25 2166 /**
0493ed11 2167 * Fix url('blah') declarations.
2168 */
2169 // $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
2170 // "url(\\1$secremoveimg\\2)", $content);
2b665f28 2171
567dc524 2172 // first check for 8bit sequences and disallowed control characters
2173 if (preg_match('/[\16-\37\200-\377]+/',$content)) {
2174 $content = '<!-- style block removed by html filter due to presence of 8bit characters -->';
2175 return array($content, $newpos);
2176 }
2177
2b665f28 2178 // IE Sucks hard. We have a special function for it.
2179 sq_fixIE_idiocy($content);
2180
2181 // remove @import line
2182 $content = preg_replace("/^\s*(@import.*)$/mi","\n<!-- @import rules forbidden -->\n",$content);
2183
5b4884be 2184 // translate ur\l and variations (IE parses that)
2b665f28 2185 // TODO check if the sq_fixIE_idiocy function already handles this.
5b4884be 2186 $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
567dc524 2187 preg_match_all("/url\s*\((.+)\)/si",$content,$aMatch);
2188 if (count($aMatch)) {
2189 $aValue = $aReplace = array();
2190 foreach($aMatch[1] as $sMatch) {
2191 // url value
2192 $urlvalue = $sMatch;
2193 sq_fix_url('style',$urlvalue, $message, $id, $mailbox,"'");
2194 $aValue[] = $sMatch;
2195 $aReplace[] = $urlvalue;
0493ed11 2196 }
567dc524 2197 $content = str_replace($aValue,$aReplace,$content);
691a2d25 2198 }
567dc524 2199
9ebace19 2200 /**
2201 * Remove any backslashes, entities, and extraneous whitespace.
2202 */
0493ed11 2203 $contentTemp = $content;
2204 sq_defang($contentTemp);
2205 sq_unspace($contentTemp);
a3daaaf3 2206
691a2d25 2207 /**
8bd0068d 2208 * Fix stupid css declarations which lead to vulnerabilities
2209 * in IE.
39352565 2210 *
2211 * Also remove "position" attribute, as it can easily be set
2212 * to "fixed" or "absolute" with "left" and "top" attributes
2213 * of zero, taking over the whole content frame. It can also
2214 * be set to relative and move itself anywhere it wants to,
2215 * displaying content in areas it shouldn't be allowed to touch.
8bd0068d 2216 */
caf0ab1d 2217 $match = Array('/\/\*.*\*\//', // removes /* blah blah */
5db90261 2218 '/expression/i',
0493ed11 2219 '/behaviou*r/i',
2220 '/binding/i',
2b665f28 2221 '/include-source/i',
2222 '/javascript/i',
39352565 2223 '/script/i',
2224 '/position/i');
2225 $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', '');
0493ed11 2226 $contentNew = preg_replace($match, $replace, $contentTemp);
2227 if ($contentNew !== $contentTemp) {
2228 // insecure css declarations are used. From now on we don't care
2229 // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
2230 $content = $contentNew;
2231 }
7d06541f 2232 return array($content, $newpos);
691a2d25 2233}
a3daaaf3 2234
0493ed11 2235
691a2d25 2236/**
8bd0068d 2237 * This function converts cid: url's into the ones that can be viewed in
2238 * the browser.
2239 *
2240 * @param $message the message object
2241 * @param $id the message id
2242 * @param $cidurl the cid: url.
2243 * @param $mailbox the message mailbox
2244 * @return a string with a http-friendly url
2245 */
b3af12ef 2246function sq_cid2http($message, $id, $cidurl, $mailbox){
691a2d25 2247 /**
8bd0068d 2248 * Get rid of quotes.
2249 */
691a2d25 2250 $quotchar = substr($cidurl, 0, 1);
2dd879b8 2251 if ($quotchar == '"' || $quotchar == "'"){
2252 $cidurl = str_replace($quotchar, "", $cidurl);
2253 } else {
2254 $quotchar = '';
2255 }
691a2d25 2256 $cidurl = substr(trim($cidurl), 4);
0493ed11 2257
2258 $match_str = '/\{.*?\}\//';
2259 $str_rep = '';
2260 $cidurl = preg_replace($match_str, $str_rep, $cidurl);
2261
e5e9381a 2262 $linkurl = find_ent_id($cidurl, $message);
9ebace19 2263 /* in case of non-safe cid links $httpurl should be replaced by a sort of
2264 unsafe link image */
e5e9381a 2265 $httpurl = '';
c8f5f606 2266
8bd0068d 2267 /**
2268 * This is part of a fix for Outlook Express 6.x generating
2269 * cid URLs without creating content-id headers. These images are
2270 * not part of the multipart/related html mail. The html contains
2271 * <img src="cid:{some_id}/image_filename.ext"> references to
2272 * attached images with as goal to render them inline although
2273 * the attachment disposition property is not inline.
2274 */
c8f5f606 2275
2276 if (empty($linkurl)) {
2277 if (preg_match('/{.*}\//', $cidurl)) {
2278 $cidurl = preg_replace('/{.*}\//','', $cidurl);
2279 if (!empty($cidurl)) {
2280 $linkurl = find_ent_id($cidurl, $message);
2281 }
2282 }
2283 }
f8a1ed5a 2284
c8f5f606 2285 if (!empty($linkurl)) {
fa26ab45 2286 $httpurl = $quotchar . sqm_baseuri() . 'src/download.php?absolute_dl=true&amp;' .
8bd0068d 2287 "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
2288 '&amp;ent_id=' . $linkurl . $quotchar;
c8f5f606 2289 } else {
2290 /**
2291 * If we couldn't generate a proper img url, drop in a blank image
2292 * instead of sending back empty, otherwise it causes unusual behaviour
2293 */
bc017c1d 2294 $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
e5e9381a 2295 }
f8a1ed5a 2296
691a2d25 2297 return $httpurl;
2298}
2299
2300/**
8bd0068d 2301 * This function changes the <body> tag into a <div> tag since we
2302 * can't really have a body-within-body.
2303 *
2304 * @param $attary an array of attributes and values of <body>
2305 * @param $mailbox mailbox we're currently reading (for cid2http)
2306 * @param $message current message (for cid2http)
2307 * @param $id current message id (for cid2http)
2308 * @return a modified array of attributes to be set for <div>
2309 */
2dd879b8 2310function sq_body2div($attary, $mailbox, $message, $id){
b583c3e8 2311 $me = 'sq_body2div';
3d8371be 2312 $divattary = Array('class' => "'bodyclass'");
b583c3e8 2313 $text = '#000000';
c189a963 2314 $has_bgc_stl = $has_txt_stl = false;
b583c3e8 2315 $styledef = '';
691a2d25 2316 if (is_array($attary) && sizeof($attary) > 0){
2317 foreach ($attary as $attname=>$attvalue){
2318 $quotchar = substr($attvalue, 0, 1);
2319 $attvalue = str_replace($quotchar, "", $attvalue);
2320 switch ($attname){
3d8371be 2321 case 'background':
8bd0068d 2322 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
3d8371be 2323 $styledef .= "background-image: url('$attvalue'); ";
2324 break;
2325 case 'bgcolor':
c189a963 2326 $has_bgc_stl = true;
3d8371be 2327 $styledef .= "background-color: $attvalue; ";
2328 break;
2329 case 'text':
c189a963 2330 $has_txt_stl = true;
3d8371be 2331 $styledef .= "color: $attvalue; ";
2332 break;
691a2d25 2333 }
a3daaaf3 2334 }
c189a963 2335 // Outlook defines a white bgcolor and no text color. This can lead to
2336 // white text on a white bg with certain themes.
2337 if ($has_bgc_stl && !$has_txt_stl) {
2338 $styledef .= "color: $text; ";
2339 }
691a2d25 2340 if (strlen($styledef) > 0){
1710ad65 2341 $divattary["style"] = "\"$styledef\"";
691a2d25 2342 }
2343 }
2344 return $divattary;
2345}
a3daaaf3 2346
691a2d25 2347/**
8bd0068d 2348 * This is the main function and the one you should actually be calling.
2349 * There are several variables you should be aware of an which need
2350 * special description.
2351 *
2352 * Since the description is quite lengthy, see it here:
2353 * http://linux.duke.edu/projects/mini/htmlfilter/
2354 *
2355 * @param $body the string with HTML you wish to filter
2356 * @param $tag_list see description above
2357 * @param $rm_tags_with_content see description above
2358 * @param $self_closing_tags see description above
2359 * @param $force_tag_closing see description above
2360 * @param $rm_attnames see description above
2361 * @param $bad_attvals see description above
2362 * @param $add_attr_to_tag see description above
2363 * @param $message message object
2364 * @param $id message id
dfc64d14 2365 * @param $recursively_called boolean flag for recursive calls into this function (optional; default FALSE)
8bd0068d 2366 * @return sanitized html safe to show on your pages.
2367 */
da2415c1 2368function sq_sanitize($body,
8bd0068d 2369 $tag_list,
2370 $rm_tags_with_content,
2371 $self_closing_tags,
2372 $force_tag_closing,
2373 $rm_attnames,
2374 $bad_attvals,
2375 $add_attr_to_tag,
2376 $message,
2377 $id,
dfc64d14 2378 $mailbox,
2379 $recursively_called=FALSE
8bd0068d 2380 ){
b583c3e8 2381 $me = 'sq_sanitize';
dfc64d14 2382
2383 /**
2384 * See if tag_list is of tags to remove or tags to allow.
2385 * false means remove these tags
2386 * true means allow these tags
2387 */
2388 $orig_tag_list = $tag_list;
7d06541f 2389 $rm_tags = array_shift($tag_list);
dfc64d14 2390
691a2d25 2391 /**
8bd0068d 2392 * Normalize rm_tags and rm_tags_with_content.
2393 */
7d06541f 2394 @array_walk($tag_list, 'sq_casenormalize');
691a2d25 2395 @array_walk($rm_tags_with_content, 'sq_casenormalize');
2396 @array_walk($self_closing_tags, 'sq_casenormalize');
dfc64d14 2397
691a2d25 2398 $curpos = 0;
2399 $open_tags = Array();
2dd879b8 2400 $trusted = "\n<!-- begin sanitized html -->\n";
691a2d25 2401 $skip_content = false;
bb8d0799 2402 /**
8bd0068d 2403 * Take care of netscape's stupid javascript entities like
2404 * &{alert('boo')};
2405 */
bb8d0799 2406 $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
691a2d25 2407
7d06541f 2408 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
691a2d25 2409 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
dfc64d14 2410
2411 /**
2412 * RCDATA and RAWTEXT tags are handled differently:
2413 * next instance of closing tag is used, whether or not
2414 * the HTML is well formed before that
2415 */
2416 global $rcdata_rawtext_tags;
2417 if (!$recursively_called
2418 && in_array($tagname, $rcdata_rawtext_tags)
2419 && $tagtype === 1){
2420 $closing_tag = false;
2421 $closing_tag_offset = $curpos;
2422 // seek out the closing tag for the current RCDATA/RAWTEXT tag
2423 while (1) {
2424 // first we need to move forward to next available closing tag
2425 // (intentionally leave off the closing > and let sq_getnxtag() validate a proper tag syntax)
2426 $next_tag = sq_findnxreg($body, $closing_tag_offset, "</\s*$tagname");
2427 if ($next_tag === false) {
2428 $closing_tag = false;
2429 break;
2430 }
2431 // but then we have to make sure it's a well-formed tag
2432 $closing_tag = sq_getnxtag($body, $next_tag[0]);
2433 if ($closing_tag === false)
2434 break;
2435 else if ($closing_tag[0] !== false
2436 // these should be redundant
2437 && $closing_tag[0] === $tagname && $closing_tag[2] === 2) {
2438 $trusted .= sq_sanitize(substr($body, $curpos, $closing_tag[4] - $curpos + 1),
2439 $orig_tag_list, $rm_tags_with_content, $self_closing_tags,
2440 $force_tag_closing, $rm_attnames, $bad_attvals, $add_attr_to_tag,
2441 $message, $id, $mailbox, true);
2442 $curpos = $closing_tag[4] + 1;
2443 continue 2;
2444 }
2445 $closing_tag_offset = $next_tag[0] + 1;
2446 }
2447 if ($closing_tag === false)
2448 { /* 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 */ }
2449 }
2450
691a2d25 2451 $free_content = substr($body, $curpos, $lt-$curpos);
2452 /**
8bd0068d 2453 * Take care of <style>
2454 */
7d06541f 2455 if ($tagname == "style" && $tagtype == 1){
da2415c1 2456 list($free_content, $curpos) =
e60a299a 2457 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
7d06541f 2458 if ($free_content != FALSE){
1d398b1d 2459 if ( !empty($attary) ) {
2460 $attary = sq_fixatts($tagname,
2461 $attary,
2462 $rm_attnames,
2463 $bad_attvals,
2464 $add_attr_to_tag,
2465 $message,
2466 $id,
2467 $mailbox
2468 );
2469 }
7d06541f 2470 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2471 $trusted .= $free_content;
2472 $trusted .= sq_tagprint($tagname, false, 2);
2473 }
2474 continue;
691a2d25 2475 }
2476 if ($skip_content == false){
2477 $trusted .= $free_content;
691a2d25 2478 }
2479 if ($tagname != FALSE){
2480 if ($tagtype == 2){
2481 if ($skip_content == $tagname){
2482 /**
8bd0068d 2483 * Got to the end of tag we needed to remove.
2484 */
691a2d25 2485 $tagname = false;
2486 $skip_content = false;
2487 } else {
2488 if ($skip_content == false){
c828931c 2489 if ($tagname == "body"){
2490 $tagname = "div";
2dd879b8 2491 }
1710ad65 2492 if (isset($open_tags[$tagname]) &&
2493 $open_tags[$tagname] > 0){
2494 $open_tags[$tagname]--;
691a2d25 2495 } else {
2dd879b8 2496 $tagname = false;
691a2d25 2497 }
691a2d25 2498 }
2499 }
2500 } else {
2501 /**
8bd0068d 2502 * $rm_tags_with_content
2503 */
691a2d25 2504 if ($skip_content == false){
2505 /**
8bd0068d 2506 * See if this is a self-closing type and change
2507 * tagtype appropriately.
2508 */
691a2d25 2509 if ($tagtype == 1
8bd0068d 2510 && in_array($tagname, $self_closing_tags)){
2dd879b8 2511 $tagtype = 3;
691a2d25 2512 }
2513 /**
8bd0068d 2514 * See if we should skip this tag and any content
2515 * inside it.
2516 */
691a2d25 2517 if ($tagtype == 1 &&
8bd0068d 2518 in_array($tagname, $rm_tags_with_content)){
691a2d25 2519 $skip_content = $tagname;
2520 } else {
da2415c1 2521 if (($rm_tags == false
8bd0068d 2522 && in_array($tagname, $tag_list)) ||
2523 ($rm_tags == true &&
2524 !in_array($tagname, $tag_list))){
691a2d25 2525 $tagname = false;
2526 } else {
2dd879b8 2527 /**
8bd0068d 2528 * Convert body into div.
2529 */
2dd879b8 2530 if ($tagname == "body"){
2531 $tagname = "div";
da2415c1 2532 $attary = sq_body2div($attary, $mailbox,
8bd0068d 2533 $message, $id);
2dd879b8 2534 }
691a2d25 2535 if ($tagtype == 1){
1710ad65 2536 if (isset($open_tags[$tagname])){
2537 $open_tags[$tagname]++;
691a2d25 2538 } else {
1710ad65 2539 $open_tags[$tagname]=1;
691a2d25 2540 }
2541 }
2542 /**
8bd0068d 2543 * This is where we run other checks.
2544 */
691a2d25 2545 if (is_array($attary) && sizeof($attary) > 0){
2546 $attary = sq_fixatts($tagname,
8bd0068d 2547 $attary,
2548 $rm_attnames,
2549 $bad_attvals,
2550 $add_attr_to_tag,
2551 $message,
2552 $id,
2553 $mailbox
2554 );
691a2d25 2555 }
2556 }
2557 }
691a2d25 2558 }
2559 }
2560 if ($tagname != false && $skip_content == false){
2561 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2562 }
691a2d25 2563 }
2564 $curpos = $gt+1;
a3daaaf3 2565 }
691a2d25 2566 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2567 if ($force_tag_closing == true){
2568 foreach ($open_tags as $tagname=>$opentimes){
2569 while ($opentimes > 0){
2570 $trusted .= '</' . $tagname . '>';
2571 $opentimes--;
2572 }
2573 }
2574 $trusted .= "\n";
2575 }
2576 $trusted .= "<!-- end sanitized html -->\n";
2577 return $trusted;
2578}
451f74a2 2579
691a2d25 2580/**
8bd0068d 2581 * This is a wrapper function to call html sanitizing routines.
2582 *
2583 * @param $body the body of the message
2584 * @param $id the id of the message
c189a963 2585
2586 * @param $message
2587 * @param $mailbox
2588 * @param boolean $take_mailto_links When TRUE, converts mailto: links
2589 * into internal SM compose links
2590 * (optional; default = TRUE)
8bd0068d 2591 * @return a string with html safe to display in the browser.
2592 */
c189a963 2593function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2594
202bcbcc 2595 // require_once(SM_PATH . 'functions/url_parser.php'); // for $MailTo_PReg_Match
c189a963 2596
691a2d25 2597 global $attachment_common_show_images, $view_unsafe_images,
dfc64d14 2598 $has_unsafe_images, $allow_svg_display, $rcdata_rawtext_tags,
2599 $remove_rcdata_rawtext_tags_and_content;
2600
2601 $rcdata_rawtext_tags = array(
2602 "noscript",
2603 "noframes",
2604 "noembed",
2605 "textarea",
2606 // also "title", "xmp", "script", "iframe", "plaintext" which we already remove below
2607 );
2608
691a2d25 2609 /**
8bd0068d 2610 * Don't display attached images in HTML mode.
2b665f28 2611 *
d0187bd6 2612 * SB: why?
8bd0068d 2613 */
691a2d25 2614 $attachment_common_show_images = false;
2615 $tag_list = Array(
dfc64d14 2616 false, // remove these tags
8bd0068d 2617 "meta",
2618 "html",
2619 "head",
2620 "base",
2621 "link",
2622 "frame",
2623 "iframe",
2624 "plaintext",
caf0ab1d 2625 "marquee",
8bd0068d 2626 );
691a2d25 2627
2628 $rm_tags_with_content = Array(
8bd0068d 2629 "script",
caf0ab1d 2630 "object",
8bd0068d 2631 "applet",
2632 "embed",
2633 "title",
2634 "frameset",
0493ed11 2635 "xmp",
caf0ab1d 2636 "xml",
8bd0068d 2637 );
c88cef21 2638 if (!$allow_svg_display)
caf0ab1d 2639 $rm_tags_with_content[] = 'svg';
dfc64d14 2640 /**
2641 * SquirrelMail will parse RCDATA and RAWTEXT tags and handle them as the special
2642 * case that they are, but if you prefer to remove them and their contents entirely
2643 * (in most cases, should be a safe thing with minimal impact), you can add the
2644 * following to config/config_local.php
2645 * $remove_rcdata_rawtext_tags_and_content = TRUE;
2646 */
2647 if ($remove_rcdata_rawtext_tags_and_content)
2648 $rm_tags_with_content = array_merge($rm_tags_with_content, $rcdata_rawtext_tags);
691a2d25 2649
2650 $self_closing_tags = Array(
8bd0068d 2651 "img",
2652 "br",
2653 "hr",
2654 "input",
caf0ab1d 2655 "outbind",
8bd0068d 2656 );
691a2d25 2657
2dd879b8 2658 $force_tag_closing = true;
691a2d25 2659
2660 $rm_attnames = Array(
8bd0068d 2661 "/.*/" =>
2662 Array(
2663 "/target/i",
2664 "/^on.*/i",
2665 "/^dynsrc/i",
2666 "/^data.*/i",
caf0ab1d 2667 "/^lowsrc.*/i",
8bd0068d 2668 )
2669 );
691a2d25 2670
6ab8321f 2671 global $use_transparent_security_image;
2672 if ($use_transparent_security_image) $secremoveimg = '../images/spacer.png';
2673 else $secremoveimg = '../images/' . _("sec_remove_eng.png");
2674
691a2d25 2675 $bad_attvals = Array(
8bd0068d 2676 "/.*/" =>
691a2d25 2677 Array(
0a6ec9b5 2678 "/^src|background/i" =>
8bd0068d 2679 Array(
691a2d25 2680 Array(
8bd0068d 2681 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2682 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2683 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
691a2d25 2684 ),
8bd0068d 2685 Array(
2686 "\\1$secremoveimg\\2",
2687 "\\1$secremoveimg\\2",
2688 "\\1$secremoveimg\\2",
8bd0068d 2689 )
2690 ),
0a6ec9b5 2691 "/^href|action/i" =>
8bd0068d 2692 Array(
0a6ec9b5 2693 Array(
8bd0068d 2694 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2695 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2696 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
0a6ec9b5 2697 ),
691a2d25 2698 Array(
8bd0068d 2699 "\\1#\\1",
2700 "\\1#\\1",
8bd0068d 2701 "\\1#\\1"
02474e43 2702 )
8bd0068d 2703 ),
2704 "/^style/i" =>
2705 Array(
2706 Array(
5db90261 2707 "/\/\*.*\*\//",
8bd0068d 2708 "/expression/i",
2709 "/binding/i",
2710 "/behaviou*r/i",
2711 "/include-source/i",
39352565 2712
2713 // position:relative can also be exploited
2714 // to put content outside of email body area
2715 // and position:fixed is similarly exploitable
2716 // as position:absolute, so we'll remove it
2717 // altogether....
2718 //
2719 // Does this screw up legitimate HTML messages?
2720 // If so, the only fix I see is to allow position
2721 // attributes (any values? I think we still have
2722 // to block static and fixed) only if $use_iframe
2723 // is enabled (1.5.0+)
2724 //
2725 // was: "/position\s*:\s*absolute/i",
2726 //
2727 "/position\s*:/i",
2728
1d935bc2 2729 "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
8bd0068d 2730 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2731 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2732 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
39352565 2733 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si",
8bd0068d 2734 ),
2735 Array(
5db90261 2736 "",
8bd0068d 2737 "idiocy",
2738 "idiocy",
2739 "idiocy",
2740 "idiocy",
567dc524 2741 "idiocy",
5b4884be 2742 "url",
8bd0068d 2743 "url(\\1#\\1)",
2744 "url(\\1#\\1)",
2745 "url(\\1#\\1)",
8bd0068d 2746 "\\1:url(\\2#\\3)"
2747 )
691a2d25 2748 )
8bd0068d 2749 )
691a2d25 2750 );
9ebace19 2751
d31f73f1 2752 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
2753 // images off by default.
afbf184c 2754 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET, FALSE);
9ebace19 2755
691a2d25 2756 if (!$view_unsafe_images){
2757 /**
8bd0068d 2758 * Remove any references to http/https if view_unsafe_images set
2759 * to false.
2760 */
1710ad65 2761 array_push($bad_attvals['/.*/']['/^src|background/i'][0],
8bd0068d 2762 '/^([\'\"])\s*https*:.*([\'\"])/si');
1710ad65 2763 array_push($bad_attvals['/.*/']['/^src|background/i'][1],
8bd0068d 2764 "\\1$secremoveimg\\1");
1710ad65 2765 array_push($bad_attvals['/.*/']['/^style/i'][0],
7ef0b415 2766 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
1710ad65 2767 array_push($bad_attvals['/.*/']['/^style/i'][1],
8bd0068d 2768 "url(\\1$secremoveimg\\1)");
691a2d25 2769 }
451f74a2 2770
691a2d25 2771 $add_attr_to_tag = Array(
8bd0068d 2772 "/^a$/i" =>
c25f2fbb 2773 Array('target'=>'"_blank"',
02474e43 2774 'title'=>'"'._("This external link will open in a new window").'"'
8bd0068d 2775 )
2776 );
da2415c1 2777 $trusted = sq_sanitize($body,
8bd0068d 2778 $tag_list,
2779 $rm_tags_with_content,
2780 $self_closing_tags,
2781 $force_tag_closing,
2782 $rm_attnames,
2783 $bad_attvals,
2784 $add_attr_to_tag,
2785 $message,
2786 $id,
2787 $mailbox
2788 );
567dc524 2789 if (strpos($trusted,$secremoveimg)){
691a2d25 2790 $has_unsafe_images = true;
da2415c1 2791 }
c189a963 2792
2793 // we want to parse mailto's in HTML output, change to SM compose links
2794 // this is a modified version of code from url_parser.php... but Marc is
2795 // right: we need a better filtering implementation; adding this randomly
2796 // here is not a great solution
2797 //
2798 if ($take_mailto_links) {
2799 // parseUrl($trusted); // this even parses URLs inside of tags... too aggressive
2800 global $MailTo_PReg_Match;
7196da62 2801 // some mailers (Microsoft, surprise surprise) produce mailto strings without being
2802 // inside an anchor (link) tag, so we have to make sure the regex looks for the
2803 // quote before mailto, and we'll also try to convert the non-links back into links
2804 $MailTo_PReg_Match = '/([\'"])?mailto:' . substr($MailTo_PReg_Match, 1) ;
c189a963 2805 if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2806 foreach ($regs[0] as $i => $mailto_before) {
7196da62 2807 $mailto_params = $regs[11][$i];
2808
2809 // get rid of any leading quote we may have captured but don't care about
2810 //
2811 $mailto_before = ltrim($mailto_before, '"\'');
2812
c189a963 2813 // get rid of any tailing quote since we have to add send_to to the end
2814 //
7196da62 2815 $mailto_before = rtrim($mailto_before, '"\'');
2816 $mailto_params = rtrim($mailto_params, '"\'');
c189a963 2817
7196da62 2818 if ($regs[2][$i]) { //if there is an email addr before '?', we need to merge it with the params
2819 $to = 'to=' . $regs[2][$i];
c189a963 2820 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
2821 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2822 else {
2823 if ($mailto_params) //already some params, append to them
2824 $mailto_params .= '&amp;' . $to;
2825 else
2826 $mailto_params .= '?' . $to;
2827 }
2828 }
2829
2830 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2831
2832 // we'll already have target=_blank, no need to allow comp_in_new
2833 // here (which would be a lot more work anyway)
2834 //
2835 global $compose_new_win;
2836 $temp_comp_in_new = $compose_new_win;
2837 $compose_new_win = 0;
2838 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2839 $compose_new_win = $temp_comp_in_new;
2840
2841 // remove <a href=" and anything after the next quote (we only
2842 // need the uri, not the link HTML) in compose uri
2843 //
7196da62 2844 // but only do this if the original mailto was in a real anchor tag
2845 //
2846 if (!empty($regs[1][$i])) {
2847 $comp_uri = substr($comp_uri, 9);
2848 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2849 }
c189a963 2850 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2851 }
2852 }
2853 }
2854
691a2d25 2855 return $trusted;
451f74a2 2856}
a4a70693 2857
da2415c1 2858/**
8bd0068d 2859 * function SendDownloadHeaders - send file to the browser
2860 *
2861 * Original Source: SM core src/download.php
2862 * moved here to make it available to other code, and separate
2863 * front end from back end functionality.
2864 *
2865 * @param string $type0 first half of mime type
2866 * @param string $type1 second half of mime type
2867 * @param string $filename filename to tell the browser for downloaded file
2868 * @param boolean $force whether to force the download dialog to pop
2869 * @param optional integer $filesize send the Content-Header and length to the browser
2870 * @return void
2871 */
02474e43 2872function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2873 global $languages, $squirrelmail_language;
cfffd60b 2874 $isIE = $isIE6plus = false;
02474e43 2875
2876 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2877
2878 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
8bd0068d 2879 strstr($HTTP_USER_AGENT, 'Opera') === false) {
cfffd60b 2880 $isIE = true;
02474e43 2881 }
2882
cfffd60b 2883 if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2884 ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2885 $isIE6plus = true;
02474e43 2886 }
2887
2888 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 2889 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
02474e43 2890 $filename =
8bd0068d 2891 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
02474e43 2892 } else {
1a4bc4a6 2893 $filename = preg_replace('/[\\\\\/:*?"<>|;]/', '_', str_replace('&nbsp;', ' ', $filename));
02474e43 2894 }
2895
2896 // A Pox on Microsoft and it's Internet Explorer!
2897 //
2898 // IE has lots of bugs with file downloads.
2899 // It also has problems with SSL. Both of these cause problems
2900 // for us in this function.
2901 //
2902 // See this article on Cache Control headers and SSL
2903 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2904 //
2905 // The best thing you can do for IE is to upgrade to the latest
2906 // version
2907 //set all the Cache Control Headers for IE
2908 if ($isIE) {
2909 $filename=rawurlencode($filename);
2910 header ("Pragma: public");
8bd0068d 2911 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
c2503a4e 2912 // does nothing - see: https://blogs.msdn.microsoft.com/ieinternals/2009/07/20/internet-explorers-cache-control-extensions/
2913 // header ("Cache-Control: post-check=0, pre-check=0", false);
8bd0068d 2914 header ("Cache-Control: private");
02474e43 2915
2916 //set the inline header for IE, we'll add the attachment header later if we need it
2917 header ("Content-Disposition: inline; filename=$filename");
2918 }
2919
2920 if (!$force) {
2921 // Try to show in browser window
2922 header ("Content-Disposition: inline; filename=\"$filename\"");
2923 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2924 } else {
2925 // Try to pop up the "save as" box
2926
2927 // IE makes this hard. It pops up 2 save boxes, or none.
2928 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2929 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2930 // But, according to Microsoft, it is "RFC compliant but doesn't
2931 // take into account some deviations that allowed within the
2932 // specification." Doesn't that mean RFC non-compliant?
2933 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2934
2935 // all browsers need the application/octet-stream header for this
2936 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2937
2938 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2939 // Do not have quotes around filename, but that applied to
2940 // "attachment"... does it apply to inline too?
2941 header ("Content-Disposition: attachment; filename=\"$filename\"");
2942
cfffd60b 2943 if ($isIE && !$isIE6plus) {
02474e43 2944 // This combination seems to work mostly. IE 5.5 SP 1 has
2945 // known issues (see the Microsoft Knowledge Base)
2946
2947 // This works for most types, but doesn't work with Word files
2948 header ("Content-Type: application/download; name=\"$filename\"");
7e2ff844 2949 header ("Content-Type: application/force-download; name=\"$filename\"");
02474e43 2950 // These are spares, just in case. :-)
2951 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2952 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2953 //header("Content-Type: application/octet-stream; name=\"$filename\"");
7e2ff844 2954 } else if ($isIE) {
2955 // This is to prevent IE for MIME sniffing and auto open a file in IE
2956 header ("Content-Type: application/force-download; name=\"$filename\"");
02474e43 2957 } else {
2958 // another application/octet-stream forces download for Netscape
2959 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2960 }
2961 }
2962
2963 //send the content-length header if the calling function provides it
2964 if ($filesize > 0) {
2965 header("Content-Length: $filesize");
2966 }
07c49f57 2967
8d863f64 2968} // end fn SendDownloadHeaders