Add handling for RCDATA and RAWTEXT elements in HTML sanitizer (CVE-2019-12970)
[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 *
8ed19238 9 * @copyright 1999-2019 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.
7a9e9c89 74 $char = strtoupper($flag{1});
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, $regs[1]);
151 /* There is some information in the content info header that could be important
8bd0068d 152 * in order to parse html messages. Let's get them here.
153 */
0600bdf1 154// if ($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 $ret;
451f74a2 191}
d4467150 192
fa26ab45 193function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding, $rStream='php://stdout', $force_crlf='') {
1035e159 194
3d8371be 195 /* Don't kill the connection if the browser is over a dialup
8bd0068d 196 * and it would take over 30 seconds to download it.
197 * Don't call set_time_limit in safe mode.
198 */
b7206e1d 199
3d8371be 200 if (!ini_get('safe_mode')) {
b7206e1d 201 set_time_limit(0);
202 }
7c7b74b3 203 /* in case of base64 encoded attachments, do not buffer them.
8bd0068d 204 Instead, echo the decoded attachment directly to screen */
7c7b74b3 205 if (strtolower($encoding) == 'base64') {
206 if (!$ent_id) {
7591f143 207 $query = "FETCH $id BODY[]";
7c7b74b3 208 } else {
7591f143 209 $query = "FETCH $id BODY[$ent_id]";
7c7b74b3 210 }
7591f143 211 sqimap_run_command($imap_stream,$query,true,$response,$message,TRUE,'sqimap_base64_decode',$rStream,true);
1d142b8d 212 } else {
7591f143 213 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
214 if (is_resource($rStream)) {
fa26ab45 215 fputs($rStream,decodeBody($body, $encoding, $force_crlf));
7591f143 216 } else {
fa26ab45 217 echo decodeBody($body, $encoding, $force_crlf);
7591f143 218 }
1d142b8d 219 }
346817d4 220
da2415c1 221 /*
8bd0068d 222 TODO, use the same method for quoted printable.
223 However, I assume that quoted printable attachments aren't that large
224 so the performancegain / memory usage drop will be minimal.
225 If we decide to add that then we need to adapt sqimap_fread because
226 we need to split te result on \n and fread doesn't stop at \n. That
227 means we also should provide $results from sqimap_fread (by ref) to
228 te function and set $no_return to false. The $filter function for
229 quoted printable should handle unsetting of $results.
230 */
da2415c1 231 /*
8bd0068d 232 TODO 2: find out how we write to the output stream php://stdout. fwrite
233 doesn't work because 'php://stdout isn't a stream.
234 */
7c7b74b3 235
5d9c6f73 236 return;
451f74a2 237}
beb9e459 238
451f74a2 239/* -[ END MIME DECODING ]----------------------------------------------------------- */
d4467150 240
3d8371be 241/* This is here for debugging purposes. It will print out a list
8bd0068d 242 * of all the entity IDs that are in the $message object.
243 */
451f74a2 244function listEntities ($message) {
3d8371be 245 if ($message) {
3c621ba1 246 echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br />';
3d8371be 247 for ($i = 0; isset($message->entities[$i]); $i++) {
248 echo "$i : ";
249 $msg = listEntities($message->entities[$i]);
250
251 if ($msg) {
252 echo "return: ";
253 return $msg;
254 }
255 }
a4a70693 256 }
451f74a2 257}
f0c4dc12 258
f792c641 259function getPriorityStr($priority) {
3d8371be 260 $priority_level = substr($priority,0,1);
261
262 switch($priority_level) {
263 /* Check for a higher then normal priority. */
264 case '1':
265 case '2':
266 $priority_string = _("High");
267 break;
268
269 /* Check for a lower then normal priority. */
270 case '4':
271 case '5':
272 $priority_string = _("Low");
273 break;
274
275 /* Check for a normal priority. */
276 case '3':
277 default:
278 $priority_level = '3';
279 $priority_string = _("Normal");
280 break;
281
282 }
283 return $priority_string;
f792c641 284}
285
451f74a2 286/* returns a $message object for a particular entity id */
287function getEntity ($message, $ent_id) {
a4a70693 288 return $message->getEntity($ent_id);
451f74a2 289}
8beafbbc 290
3d8371be 291/* translateText
8bd0068d 292 * Extracted from strings.php 23/03/2002
293 */
da4c66e8 294
295function translateText(&$body, $wrap_at, $charset) {
3d8371be 296 global $where, $what; /* from searching */
297 global $color; /* color theme */
da4c66e8 298
202bcbcc 299 // require_once(SM_PATH . 'functions/url_parser.php');
da4c66e8 300
301 $body_ary = explode("\n", $body);
da4c66e8 302 for ($i=0; $i < count($body_ary); $i++) {
4c1f4be3 303 $line = rtrim($body_ary[$i],"\r");
304
da4c66e8 305 if (strlen($line) - 2 >= $wrap_at) {
c7aff938 306 sqWordWrap($line, $wrap_at, $charset);
da4c66e8 307 }
308 $line = charset_decode($charset, $line);
309 $line = str_replace("\t", ' ', $line);
310
311 parseUrl ($line);
312
3d8371be 313 $quotes = 0;
da4c66e8 314 $pos = 0;
3d8371be 315 $j = strlen($line);
da4c66e8 316
3d8371be 317 while ($pos < $j) {
da4c66e8 318 if ($line[$pos] == ' ') {
3d8371be 319 $pos++;
da4c66e8 320 } else if (strpos($line, '&gt;', $pos) === $pos) {
321 $pos += 4;
3d8371be 322 $quotes++;
da4c66e8 323 } else {
324 break;
325 }
326 }
3d8371be 327
83c94382 328 if ($quotes % 2) {
d0814c02 329 $line = '<span class="quote1">' . $line . '</span>';
4c25967c 330 } elseif ($quotes) {
d0814c02 331 $line = '<span class="quote2">' . $line . '</span>';
da4c66e8 332 }
3d8371be 333
da4c66e8 334 $body_ary[$i] = $line;
335 }
336 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
337}
338
a2bfcbce 339/**
da1b55ad 340 * This returns a parsed string called $body. That string can then
341 * be displayed as the actual message in the HTML. It contains
342 * everything needed, including HTML Tags, Attachments at the
343 * bottom, etc.
f8a1ed5a 344 *
da1b55ad 345 * Since 1.2.0 function uses message_body hook.
346 * Till 1.3.0 function included output of formatAttachments().
347 *
348 * @param resource $imap_stream imap connection resource
349 * @param object $message squirrelmail message object
350 * @param array $color squirrelmail color theme array
351 * @param integer $wrap_at number of characters per line
352 * @param string $ent_num (since 1.3.0) message part id
353 * @param integer $id (since 1.3.0) message id
354 * @param string $mailbox (since 1.3.0) imap folder name
da1b55ad 355 * @return string html formated message text
356 */
a540f994 357function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
3d8371be 358 /* This if statement checks for the entity to show as the
8bd0068d 359 * primary message. To add more of them, just put them in the
360 * order that is their priority.
361 */
ce68b76b 362 global $startMessage, $languages, $squirrelmail_language,
40a34e57 363 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id,
14c85e39 364 $use_iframe, $iframe_height, $download_and_unsafe_link,
955bfc8f 365 $download_href, $unsafe_image_toggle_href, $unsafe_image_toggle_text,
551c7b53 366 $oTemplate, $nbsp;
2c25d36a 367
368 // workaround for not updated config.php
369 if (! isset($use_iframe)) $use_iframe = false;
77bfbd2e 370
d31f73f1 371 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
372 // images off by default.
afbf184c 373 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET, FALSE);
d03c24f4 374
cc34b00d 375 $body = '';
23bcec6f 376 $urlmailbox = urlencode($mailbox);
451f74a2 377 $body_message = getEntity($message, $ent_num);
378 if (($body_message->header->type0 == 'text') ||
8bd0068d 379 ($body_message->header->type0 == 'rfc822')) {
3d8371be 380 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
451f74a2 381 $body = decodeBody($body, $body_message->header->encoding);
e842b215 382
383 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 384 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
e842b215 385 if (mb_detect_encoding($body) != 'ASCII') {
33a55f5a 386 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
e842b215 387 }
388 }
d849b570 389
390 /* As of 1.5.2, $body is passed (and modified) by reference */
391 do_hook('message_body', $body);
23bcec6f 392
3d8371be 393 /* If there are other types that shouldn't be formatted, add
8bd0068d 394 * them here.
395 */
3d8371be 396
451f74a2 397 if ($body_message->header->type1 == 'html') {
3d8371be 398 if ($show_html_default <> 1) {
85015544 399 $entity_conv = array('&nbsp;' => ' ',
8bd0068d 400 '<p>' => "\n",
401 '<P>' => "\n",
402 '<br>' => "\n",
403 '<BR>' => "\n",
404 '<br />' => "\n",
405 '<BR />' => "\n",
406 '&gt;' => '>',
407 '&lt;' => '<');
85015544 408 $body = strtr($body, $entity_conv);
3d8371be 409 $body = strip_tags($body);
85015544 410 $body = trim($body);
411 translateText($body, $wrap_at,
8bd0068d 412 $body_message->header->getParameter('charset'));
a540f994 413 } elseif ($use_iframe) {
84410f31 414 /**
415 * If we don't add html message between iframe tags,
416 * we must detect unsafe images and modify $has_unsafe_images.
f8a1ed5a 417 */
758a7889 418 $html_body = magicHTML($body, $id, $message, $mailbox);
b6c52e61 419 // Convert character set in order to display html mails in different character set
420 $html_body = charset_decode($body_message->header->getParameter('charset'),$html_body,false,true);
84410f31 421
2c25d36a 422 // creating iframe url
423 $iframeurl=sqm_baseuri().'src/view_html.php?'
f8a1ed5a 424 . 'mailbox=' . $urlmailbox
2c25d36a 425 . '&amp;passed_id=' . $id
426 . '&amp;ent_id=' . $ent_num
427 . '&amp;view_unsafe_images=' . (int) $view_unsafe_images;
428
79d58d4c 429 global $oTemplate;
430 $oTemplate->assign('iframe_url', $iframeurl);
51554708 431 $oTemplate->assign('iframe_height', $iframe_height);
79d58d4c 432 $oTemplate->assign('html_body', $html_body);
2b665f28 433
79d58d4c 434 $body = $oTemplate->fetch('read_html_iframe.tpl');
a3daaaf3 435 } else {
2c25d36a 436 // old way of html rendering
b6c52e61 437 /**
758a7889 438 * convert character set. charset_decode does not remove html special chars
b6c52e61 439 * applied by magicHTML functions and does not sanitize them second time if
758a7889 440 * fourth argument is true.
441 */
567dc524 442 $charset = $body_message->header->getParameter('charset');
443 if (!empty($charset)) {
444 $body = charset_decode($charset,$body,false,true);
445 }
446 $body = magicHTML($body, $id, $message, $mailbox);
a3daaaf3 447 }
451f74a2 448 } else {
3d8371be 449 translateText($body, $wrap_at,
8bd0068d 450 $body_message->header->getParameter('charset'));
451f74a2 451 }
a2bfcbce 452
9ebace19 453 /*
454 * Previously the links for downloading and unsafe images were printed
455 * under the mail. By putting the links in a global variable we can
456 * print it in the toolbar where it belongs. Since the original code was
457 * in this place it's left here. It might be possible to move it to some
458 * other place if that makes sense. The possibility to do so has not
459 * been evaluated yet.
460 */
461
462 // Initialize the global variable to an empty string.
463 // FIXME: To have $download_and_unsafe_link as a global variable might not be needed since the use of separate variables ($download_href, $unsafe_image_toggle_href, and $unsafe_image_toggle_text) for the templates was introduced.
40a34e57 464 $download_and_unsafe_link = '';
465
9ebace19 466 // Prepare and build a link for downloading the mail.
83cf04bd 467 $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
8bd0068d 468 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
469 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
08b7f7cc 470 if (isset($passed_ent_id)) {
471 $link .= '&amp;passed_ent_id='.$passed_ent_id;
472 }
14c85e39 473 $download_href = SM_PATH . 'src/download.php?absolute_dl=true&amp;' . $link;
9ebace19 474
475 // Always add the link for downloading the mail as a file to the global
476 // variable.
955bfc8f 477 $download_and_unsafe_link .= "$nbsp|$nbsp"
478 . create_hyperlink($download_href, _("Download this as a file"));
9ebace19 479
480 // Find out the right text to use in the link depending on the
481 // circumstances. If the unsafe images are displayed the link should
482 // hide them, if they aren't displayed the link should only appear if
483 // the mail really contains unsafe images.
7aad7b77 484 if ($view_unsafe_images) {
23f617b8 485 $text = _("Hide Unsafe Images");
7aad7b77 486 } else {
08b7f7cc 487 if (isset($has_unsafe_images) && $has_unsafe_images) {
488 $link .= '&amp;view_unsafe_images=1';
489 $text = _("View Unsafe Images");
490 } else {
491 $text = '';
492 }
3d8371be 493 }
9ebace19 494
495 // Only create a link for unsafe images if there's need for one. If so:
496 // add it to the global variable.
83cf04bd 497 if($text != '') {
14c85e39 498 $unsafe_image_toggle_href = SM_PATH . 'src/read_body.php?'.$link;
499 $unsafe_image_toggle_text = $text;
955bfc8f 500 $download_and_unsafe_link .= "$nbsp|$nbsp"
501 . create_hyperlink($unsafe_image_toggle_href, $text);
83cf04bd 502 }
3d8371be 503 }
504 return $body;
451f74a2 505}
b74ba498 506
da1b55ad 507/**
a540f994 508 * Generate attachments array for passing to templates.
2b665f28 509 *
d67f519a 510 * @since 1.5.2
da1b55ad 511 * @param object $message SquirrelMail message object
512 * @param array $exclude_id message parts that are not attachments.
513 * @param string $mailbox mailbox name
514 * @param integer $id message id
da1b55ad 515 */
d67f519a 516function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
caf0ab1d 517 global $where, $what, $startMessage, $color, $passed_ent_id,
518 $base_uri, $block_svg_download;
451f74a2 519
23bcec6f 520 $att_ar = $message->getAttachments($exclude_id);
23bcec6f 521 $urlMailbox = urlencode($mailbox);
522
d67f519a 523 $attachments = array();
23bcec6f 524 foreach ($att_ar as $att) {
fdc9d9b5 525 $ent = $att->entity_id;
2e25760a 526 $header = $att->header;
f0c4dc12 527 $type0 = strtolower($header->type0);
528 $type1 = strtolower($header->type1);
caf0ab1d 529 if ($block_svg_download && strpos($type1, 'svg') === 0)
530 continue;
531
2e25760a 532 $name = '';
d0187bd6 533 $links = array();
21dab2dc 534 $links['download link']['text'] = _("Download");
202bcbcc 535 $links['download link']['href'] = $base_uri .
8bd0068d 536 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
2b665f28 537
2e25760a 538 if ($type0 =='message' && $type1 == 'rfc822') {
202bcbcc 539 $default_page = $base_uri . 'src/read_body.php';
2e25760a 540 $rfc822_header = $att->rfc822_header;
098ea084 541 $filename = $rfc822_header->subject;
6cc08d8b 542 if (trim( $filename ) == '') {
543 $filename = 'untitled-[' . $ent . ']' ;
08b7f7cc 544 }
2e25760a 545 $from_o = $rfc822_header->from;
546 if (is_object($from_o)) {
04ea844e 547 $from_name = decodeHeader($from_o->getAddress(false));
7e697748 548 } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
549 // something weird happens when a digest message is opened and you return to the digest
550 // now the from object is part of an array. Probably the parseHeader call overwrites the info
551 // retrieved from the bodystructure in a different way. We need to fix this later.
552 // possible starting point, do not fetch header we already have and inspect how
553 // the rfc822_header object behaves.
554 $from_name = decodeHeader($from_o[0]->getAddress(false));
2e25760a 555 } else {
556 $from_name = _("Unknown sender");
f0c4dc12 557 }
d0187bd6 558 $description = _("From").': '.$from_name;
23bcec6f 559 } else {
202bcbcc 560 $default_page = $base_uri . 'src/download.php';
02474e43 561 $filename = $att->getFilename();
f810c0b2 562 if ($header->description) {
098ea084 563 $description = decodeHeader($header->description);
f810c0b2 564 } else {
3d8371be 565 $description = '';
566 }
2e25760a 567 }
568
569 $display_filename = $filename;
3d8371be 570 if (isset($passed_ent_id)) {
571 $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
572 } else {
573 $passed_ent_id_link = '';
574 }
575 $defaultlink = $default_page . "?startMessage=$startMessage"
8bd0068d 576 . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
577 . '&amp;ent_id='.$ent.$passed_ent_id_link;
2e25760a 578 if ($where && $what) {
8bd0068d 579 $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
2e25760a 580 }
7e2ff844 581 // IE does make use of mime content sniffing. Forcing a download
582 // prohibit execution of XSS inside an application/octet-stream attachment
583 if ($type0 == 'application' && $type1 == 'octet-stream') {
584 $defaultlink .= '&amp;absolute_dl=true';
585 }
2b665f28 586
3d8371be 587 /* This executes the attachment hook with a specific MIME-type.
53901c7b 588 * It also allows plugins to run if there's a rule for a more
589 * generic type. Finally, a hook for ALL attachment types is
590 * run as well.
8bd0068d 591 */
8dca4d22 592 // First remember the default link.
593 $defaultlink_orig = $defaultlink;
594
d849b570 595 /* The API for this hook has changed as of 1.5.2 so that all plugin
596 arguments are passed in an array instead of each their own plugin
597 argument, and arguments are passed by reference, so instead of
598 returning any changes, changes should simply be made to the original
599 arguments themselves. */
53901c7b 600 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
9c3b2d22 601 &$defaultlink, &$display_filename, &$where, &$what);
602 do_hook("attachment $type0/$type1", $temp);
53901c7b 603 /* The API for this hook has changed as of 1.5.2 so that all plugin
604 arguments are passed in an array instead of each their own plugin
605 argument, and arguments are passed by reference, so instead of
606 returning any changes, changes should simply be made to the original
607 arguments themselves. */
608 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
609 &$defaultlink, &$display_filename, &$where, &$what);
610 // Do not let a generic plugin change the default link if a more
611 // specialized one already did it...
612 if ($defaultlink != $defaultlink_orig) {
613 $dummy = '';
614 $temp[5] = &$dummy;
2e25760a 615 }
53901c7b 616 do_hook("attachment $type0/*", $temp);
d849b570 617 /* The API for this hook has changed as of 1.5.2 so that all plugin
618 arguments are passed in an array instead of each their own plugin
619 argument, and arguments are passed by reference, so instead of
620 returning any changes, changes should simply be made to the original
621 arguments themselves. */
9c3b2d22 622 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
623 &$defaultlink, &$display_filename, &$where, &$what);
8dca4d22 624 // Do not let a generic plugin change the default link if a more
625 // specialized one already did it...
626 if ($defaultlink != $defaultlink_orig) {
627 $dummy = '';
628 $temp[5] = &$dummy;
629 }
9c3b2d22 630 do_hook("attachment */*", $temp);
77b88425 631
d67f519a 632 $this_attachment = array();
633 $this_attachment['Name'] = decodeHeader($display_filename);
634 $this_attachment['Description'] = $description;
635 $this_attachment['DefaultHREF'] = $defaultlink;
636 $this_attachment['DownloadHREF'] = $links['download link']['href'];
637 $this_attachment['ViewHREF'] = isset($links['attachment_common']) ? $links['attachment_common']['href'] : '';
638 $this_attachment['Size'] = $header->size;
3047e291 639 $this_attachment['ContentType'] = sm_encode_html_special_chars($type0 .'/'. $type1);
d67f519a 640 $this_attachment['OtherLinks'] = array();
3d8371be 641 foreach ($links as $val) {
d0187bd6 642 if ($val['text']==_("Download") || $val['text'] == _("View"))
643 continue;
644 if (empty($val['text']) && empty($val['extra']))
645 continue;
2b665f28 646
d67f519a 647 $temp = array();
648 $temp['HREF'] = $val['href'];
649 $temp['Text'] = (empty($val['text']) ? '' : $val['text']) . (empty($val['extra']) ? '' : $val['extra']);
650 $this_attachment['OtherLinks'][] = $temp;
2e25760a 651 }
d67f519a 652 $attachments[] = $this_attachment;
2b665f28 653
3d8371be 654 unset($links);
2e25760a 655 }
2b665f28 656
d67f519a 657 return $attachments;
658}
659
660/**
661 * Displays attachment links and information
662 *
663 * Since 1.3.0 function is not included in formatBody() call.
664 *
665 * Since 1.0.2 uses attachment $type0/$type1 hook.
666 * Since 1.2.5 uses attachment $type0/* hook.
667 * Since 1.5.0 uses attachments_bottom hook.
668 * Since 1.5.2 uses templates and does *not* return a value.
669 *
670 * @param object $message SquirrelMail message object
671 * @param array $exclude_id message parts that are not attachments.
672 * @param string $mailbox mailbox name
673 * @param integer $id message id
674 */
675function formatAttachments($message, $exclude_id, $mailbox, $id) {
676 global $oTemplate;
2b665f28 677
d67f519a 678 $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
d0187bd6 679
680 $oTemplate->assign('attachments', $attach);
681 $oTemplate->display('read_attachments.tpl');
451f74a2 682}
b74ba498 683
7c7b74b3 684function sqimap_base64_decode(&$string) {
7c0ec1d8 685
b17a8968 686 // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
7c0ec1d8 687 // fly decoding (to reduce memory usage) you have to check if the
688 // data has incomplete pairs
689
b17a8968 690 // Remove the noise in order to check if the 4 bytes pairs are complete
7c0ec1d8 691 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
692
42ce44f8 693 $sStringRem = '';
7c0ec1d8 694 $iMod = strlen($string) % 4;
695 if ($iMod) {
696 $sStringRem = substr($string,-$iMod);
b17a8968 697 // Check if $sStringRem contains padding characters
7c0ec1d8 698 if (substr($sStringRem,-1) != '=') {
699 $string = substr($string,0,-$iMod);
700 } else {
701 $sStringRem = '';
702 }
703 }
7c7b74b3 704 $string = base64_decode($string);
7c0ec1d8 705 return $sStringRem;
7c7b74b3 706}
707
fdf7cef1 708/**
fa26ab45 709 * Decodes encoded string (usually message body)
710 *
711 * This function decodes a string (usually the message body)
712 * depending on the encoding type. Currently quoted-printable
713 * and base64 encodings are supported.
714 *
715 * The decode_body hook was added to this function in 1.4.2/1.5.0.
716 * The $force_crlf parameter was added in 1.5.2.
717 *
718 * @param string $string The encoded string
719 * @param string $encoding used encoding
720 * @param string $force_crlf Whether or not to force CRLF or LF
721 * line endings (or to leave as is).
722 * If given as "LF", line endings will
723 * all be converted to LF; if "CRLF",
724 * line endings will all be converted
725 * to CRLF. If given as an empty value,
9785376e 726 * the global $force_crlf_default will
fa26ab45 727 * be consulted (it can be specified in
728 * config/config_local.php). Otherwise,
729 * any other value will cause the string
730 * to be left alone. Note that this will
731 * be overridden to "LF" if not using at
732 * least PHP version 4.3.0. (OPTIONAL;
733 * default is empty - consult global
734 * default value)
735 *
736 * @return string The decoded string
fdf7cef1 737 *
fdf7cef1 738 * @since 1.0
fa26ab45 739 *
fdf7cef1 740 */
fa26ab45 741function decodeBody($string, $encoding, $force_crlf='') {
742
743 global $force_crlf_default;
744 if (empty($force_crlf)) $force_crlf = $force_crlf_default;
745 $force_crlf = strtoupper($force_crlf);
746
747 // must force line endings to LF due to broken
748 // quoted_printable_decode() in PHP versions
749 // before 4.3.0 (see below)
750 //
751 if (!check_php_version(4, 3, 0) || $force_crlf == 'LF')
752 $string = str_replace("\r\n", "\n", $string);
753 else if ($force_crlf == 'CRLF')
754 $string = str_replace("\n", "\r\n", $string);
83be314a 755
b583c3e8 756 $encoding = strtolower($encoding);
3d8371be 757
d849b570 758 $encoding_handler = do_hook('decode_body', $encoding);
5166f86a 759
760
fa26ab45 761 // plugins get first shot at decoding the string
5166f86a 762 //
763 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
fa26ab45 764 $string = $encoding_handler('decode', $string);
5166f86a 765
fdf7cef1 766 } elseif ($encoding == 'quoted-printable' ||
8bd0068d 767 $encoding == 'quoted_printable') {
fa26ab45 768
769 // quoted_printable_decode() function is broken in older
770 // php versions. Text with \r\n decoding was fixed only
771 // in php 4.3.0. Minimal code requirement is PHP 4.0.4+
772 // and the above call to: str_replace("\r\n", "\n", $string);
773 //
774 $string = quoted_printable_decode($string);
775
fdf7cef1 776 } elseif ($encoding == 'base64') {
fa26ab45 777 $string = base64_decode($string);
b583c3e8 778 }
3d8371be 779
b583c3e8 780 // All other encodings are returned raw.
fa26ab45 781 return $string;
451f74a2 782}
783
9f7f68c3 784/**
8bd0068d 785 * Decodes headers
786 *
e1115979 787 * This function decodes strings that are encoded according to
8bd0068d 788 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
789 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
790 *
791 * @param string $string header string that has to be made readable
792 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
bcae324b 793 * @param boolean $htmlsafe preserve spaces and sanitize html special characters. defaults to true
8bd0068d 794 * @param boolean $decide decide if string can be utfencoded. defaults to false
795 * @return string decoded header string
796 */
bcae324b 797function decodeHeader ($string, $utfencode=true,$htmlsafe=true,$decide=false) {
caf0ab1d 798 global $languages, $squirrelmail_language,$default_charset, $fix_broken_base64_encoded_messages;
79e07c7e 799 if (is_array($string)) {
800 $string = implode("\n", $string);
801 }
da2415c1 802
10dec454 803 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 804 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
33a55f5a 805 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
08b7f7cc 806 // Do we need to return at this point?
807 // return $string;
83be314a 808 }
79e07c7e 809 $i = 0;
08b7f7cc 810 $iLastMatch = -2;
db65b6b0 811 $encoded = true;
0a06275a 812
821d1f14 813// FIXME: spaces are allowed inside quoted-printable encoding, but the following line will bust up any such encoded strings
098ea084 814 $aString = explode(' ',$string);
08b7f7cc 815 $ret = '';
098ea084 816 foreach ($aString as $chunk) {
358a78a1 817 if ($encoded && $chunk === '') {
08b7f7cc 818 continue;
358a78a1 819 } elseif ($chunk === '') {
08b7f7cc 820 $ret .= ' ';
821 continue;
822 }
098ea084 823 $encoded = false;
08b7f7cc 824 /* if encoded words are not separated by a linear-space-white we still catch them */
825 $j = $i-1;
7e6ca3e8 826
08b7f7cc 827 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
828 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
829 if ($iLastMatch !== $j) {
bcae324b 830 if ($htmlsafe) {
9f7f68c3 831 $ret .= '&#32;';
08b7f7cc 832 } else {
833 $ret .= ' ';
834 }
835 }
836 $iLastMatch = $i;
837 $j = $i;
bcae324b 838 if ($htmlsafe) {
3047e291 839 $ret .= sm_encode_html_special_chars($res[1]);
cb718de0 840 } else {
841 $ret .= $res[1];
842 }
098ea084 843 $encoding = ucfirst($res[3]);
d6f584fc 844
845 /* decide about valid decoding */
846 if ($decide && is_conversion_safe($res[2])) {
8bd0068d 847 $utfencode=true;
848 $can_be_encoded=true;
d6f584fc 849 } else {
8bd0068d 850 $can_be_encoded=false;
d6f584fc 851 }
098ea084 852 switch ($encoding)
853 {
8bd0068d 854 case 'B':
caf0ab1d 855 // fix broken base64-encoded strings (remove end = padding,
856 // change any = to + in middle of string, add padding back
857 // to the end)
858 if ($fix_broken_base64_encoded_messages) {
859 $encoded_string_minus_padding = strtr(rtrim($res[4], '='), '=', '+');
860 $res[4] = str_pad($encoded_string_minus_padding, strlen($res[4]), '=');
861 }
8bd0068d 862 $replace = base64_decode($res[4]);
863 if ($utfencode) {
864 if ($can_be_encoded) {
865 /* convert string to different charset,
866 * if functions asks for it (usually in compose)
867 */
bcae324b 868 $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsafe);
8bd0068d 869 } else {
870 // convert string to html codes in order to display it
871 $ret .= charset_decode($res[2],$replace);
872 }
fab65ca9 873 } else {
bcae324b 874 if ($htmlsafe) {
3047e291 875 $replace = sm_encode_html_special_chars($replace);
8bd0068d 876 }
877 $ret.= $replace;
fab65ca9 878 }
8bd0068d 879 break;
880 case 'Q':
881 $replace = str_replace('_', ' ', $res[4]);
6c4e2e9b 882 $replace = preg_replace_callback('/=([0-9a-f]{2})/i',
883 create_function ('$matches', 'return chr(hexdec($matches[1]));'),
8bd0068d 884 $replace);
885 if ($utfencode) {
886 if ($can_be_encoded) {
887 /* convert string to different charset,
888 * if functions asks for it (usually in compose)
889 */
bcae324b 890 $replace = charset_convert($res[2], $replace,$default_charset,$htmlsafe);
8bd0068d 891 } else {
892 // convert string to html codes in order to display it
893 $replace = charset_decode($res[2], $replace);
894 }
895 } else {
bcae324b 896 if ($htmlsafe) {
3047e291 897 $replace = sm_encode_html_special_chars($replace);
8bd0068d 898 }
098ea084 899 }
8bd0068d 900 $ret .= $replace;
901 break;
902 default:
903 break;
79e07c7e 904 }
098ea084 905 $chunk = $res[5];
906 $encoded = true;
08b7f7cc 907 }
908 if (!$encoded) {
bcae324b 909 if ($htmlsafe) {
9f7f68c3 910 $ret .= '&#32;';
08b7f7cc 911 } else {
912 $ret .= ' ';
da2415c1 913 }
08b7f7cc 914 }
dc3d13a7 915
bcae324b 916 if (!$encoded && $htmlsafe) {
3047e291 917 $ret .= sm_encode_html_special_chars($chunk);
dc3d13a7 918 } else {
919 $ret .= $chunk;
920 }
098ea084 921 ++$i;
922 }
fd81e884 923 /* remove the first added space */
924 if ($ret) {
bcae324b 925 if ($htmlsafe) {
9f7f68c3 926 $ret = substr($ret,5);
fd81e884 927 } else {
928 $ret = substr($ret,1);
929 }
930 }
da2415c1 931
08b7f7cc 932 return $ret;
451f74a2 933}
934
9f7f68c3 935/**
a24cf710 936 * Encodes header
8bd0068d 937 *
a24cf710 938 * Function uses XTRA_CODE _encodeheader function, if such function exists.
a24cf710 939 *
758a7889 940 * Function uses Q encoding by default and encodes a string according to RFC
941 * 1522 for use in headers if it contains 8-bit characters or anything that
a24cf710 942 * looks like it should be encoded.
8bd0068d 943 *
758a7889 944 * Function switches to B encoding and encodeHeaderBase64() function, if
945 * string is 8bit and multibyte character set supported by mbstring extension
946 * is used. It can cause E_USER_NOTICE errors, if interface is used with
f270a6eb 947 * multibyte character set unsupported by mbstring extension.
948 *
8bd0068d 949 * @param string $string header string, that has to be encoded
950 * @return string quoted-printable encoded string
f270a6eb 951 * @todo make $mb_charsets system wide constant
8bd0068d 952 */
451f74a2 953function encodeHeader ($string) {
6fbd125b 954 global $default_charset, $languages, $squirrelmail_language;
83be314a 955
10dec454 956 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 957 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
33a55f5a 958 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
83be314a 959 }
793cc001 960
a24cf710 961 // Use B encoding for multibyte charsets
f270a6eb 962 $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
963 if (in_array($default_charset,$mb_charsets) &&
964 in_array($default_charset,sq_mb_list_encodings()) &&
965 sq_is8bit($string)) {
966 return encodeHeaderBase64($string,$default_charset);
967 } elseif (in_array($default_charset,$mb_charsets) &&
968 sq_is8bit($string) &&
969 ! in_array($default_charset,sq_mb_list_encodings())) {
970 // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
971 // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
972 }
a24cf710 973
451f74a2 974 // Encode only if the string contains 8-bit characters or =?
3d8371be 975 $j = strlen($string);
098ea084 976 $max_l = 75 - strlen($default_charset) - 7;
977 $aRet = array();
451f74a2 978 $ret = '';
c96c32f4 979 $iEncStart = $enc_init = false;
0d53f0f9 980 $cur_l = $iOffset = 0;
3d8371be 981 for($i = 0; $i < $j; ++$i) {
c96c32f4 982 switch($string{$i})
983 {
6b76cffa 984 case '"':
8bd0068d 985 case '=':
986 case '<':
987 case '>':
988 case ',':
989 case '?':
990 case '_':
991 if ($iEncStart === false) {
992 $iEncStart = $i;
993 }
994 $cur_l+=3;
995 if ($cur_l > ($max_l-2)) {
996 /* if there is an stringpart that doesn't need encoding, add it */
08b7f7cc 997 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
c96c32f4 998 $aRet[] = "=?$default_charset?Q?$ret?=";
999 $iOffset = $i;
1000 $cur_l = 0;
1001 $ret = '';
1002 $iEncStart = false;
08b7f7cc 1003 } else {
8bd0068d 1004 $ret .= sprintf("=%02X",ord($string{$i}));
c96c32f4 1005 }
8bd0068d 1006 break;
1007 case '(':
1008 case ')':
1009 if ($iEncStart !== false) {
08b7f7cc 1010 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
8bd0068d 1011 $aRet[] = "=?$default_charset?Q?$ret?=";
c96c32f4 1012 $iOffset = $i;
8bd0068d 1013 $cur_l = 0;
1014 $ret = '';
1015 $iEncStart = false;
c96c32f4 1016 }
8bd0068d 1017 break;
1018 case ' ':
c96c32f4 1019 if ($iEncStart !== false) {
098ea084 1020 $cur_l++;
1021 if ($cur_l > $max_l) {
08b7f7cc 1022 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
c96c32f4 1023 $aRet[] = "=?$default_charset?Q?$ret?=";
c96c32f4 1024 $iOffset = $i;
1025 $cur_l = 0;
098ea084 1026 $ret = '';
8bd0068d 1027 $iEncStart = false;
08b7f7cc 1028 } else {
8bd0068d 1029 $ret .= '_';
c96c32f4 1030 }
3d8371be 1031 }
8bd0068d 1032 break;
1033 default:
1034 $k = ord($string{$i});
1035 if ($k > 126) {
1036 if ($iEncStart === false) {
1037 // do not start encoding in the middle of a string, also take the rest of the word.
1038 $sLeadString = substr($string,0,$i);
1039 $aLeadString = explode(' ',$sLeadString);
1040 $sToBeEncoded = array_pop($aLeadString);
1041 $iEncStart = $i - strlen($sToBeEncoded);
1042 $ret .= $sToBeEncoded;
1043 $cur_l += strlen($sToBeEncoded);
1044 }
1045 $cur_l += 3;
1046 /* first we add the encoded string that reached it's max size */
1047 if ($cur_l > ($max_l-2)) {
1048 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1049 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
1050 $cur_l = 3;
1051 $ret = '';
1052 $iOffset = $i;
1053 $iEncStart = $i;
1054 }
1055 $enc_init = true;
1056 $ret .= sprintf("=%02X", $k);
1057 } else {
1058 if ($iEncStart !== false) {
1059 $cur_l++;
1060 if ($cur_l > $max_l) {
1061 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1062 $aRet[] = "=?$default_charset?Q?$ret?=";
1063 $iEncStart = false;
1064 $iOffset = $i;
1065 $cur_l = 0;
1066 $ret = '';
1067 } else {
1068 $ret .= $string{$i};
1069 }
1070 }
1071 }
1072 break;
f7b3ba37 1073 }
451f74a2 1074 }
793cc001 1075
c96c32f4 1076 if ($enc_init) {
1077 if ($iEncStart !== false) {
1078 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1079 $aRet[] = "=?$default_charset?Q?$ret?=";
1080 } else {
1081 $aRet[] = substr($string,$iOffset);
1082 }
1083 $string = implode('',$aRet);
451f74a2 1084 }
3d8371be 1085 return $string;
451f74a2 1086}
b74ba498 1087
f270a6eb 1088/**
1089 * Encodes string according to rfc2047 B encoding header formating rules
1090 *
758a7889 1091 * It is recommended way to encode headers with character sets that store
f270a6eb 1092 * symbols in more than one byte.
1093 *
1094 * Function requires mbstring support. If required mbstring functions are missing,
1095 * function returns false and sets E_USER_WARNING level error message.
1096 *
758a7889 1097 * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
1098 * that mbstring functions will generate E_WARNING errors, if unsupported
f270a6eb 1099 * character set is used. mb_encode_mimeheader function provided by php
1100 * mbstring extension is not used in order to get better control of header
1101 * encoding.
1102 *
758a7889 1103 * Used php code functions - function_exists(), trigger_error(), strlen()
1104 * (is used with charset names and base64 strings). Used php mbstring
f270a6eb 1105 * functions - mb_strlen and mb_substr.
1106 *
758a7889 1107 * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
f270a6eb 1108 * encoding), rfc 2822 (header folding)
1109 *
1110 * @param string $string header string that must be encoded
758a7889 1111 * @param string $charset character set. Must be supported by mbstring extension.
f270a6eb 1112 * Use sq_mb_list_encodings() to detect supported charsets.
1113 * @return string string encoded according to rfc2047 B encoding formating rules
1114 * @since 1.5.1
1115 * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1116 * @todo Do we want to control max length of header?
1117 * @todo Do we want to control EOL (end-of-line) marker?
1118 * @todo Do we want to translate error message?
1119 */
1120function encodeHeaderBase64($string,$charset) {
1121 /**
1122 * Check mbstring function requirements.
1123 */
1124 if (! function_exists('mb_strlen') ||
1125 ! function_exists('mb_substr')) {
1126 // set E_USER_WARNING
1127 trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
1128 // return false
1129 return false;
1130 }
1131
1132 // initial return array
1133 $aRet = array();
1134
1135 /**
1136 * header length = 75 symbols max (same as in encodeHeader)
1137 * remove $charset length
1138 * remove =? ? ?= (5 chars)
1139 * remove 2 more chars (\r\n ?)
1140 */
1141 $iMaxLength = 75 - strlen($charset) - 7;
1142
1143 // set first character position
1144 $iStartCharNum = 0;
1145
1146 // loop through all characters. count characters and not bytes.
1147 for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1148 // encode string from starting character to current character.
1149 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1150
1151 // Check encoded string length
1152 if(strlen($encoded_string)>$iMaxLength) {
1153 // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1154 $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1155
1156 // set new starting character
1157 $iStartCharNum = $iCharNum-1;
1158
1159 // encode last char (in case it is last character in string)
1160 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1161 } // if string is shorter than max length - add next character
1162 }
1163
1164 // add last encoded string to array
1165 $aRet[] = $encoded_string;
1166
1167 // set initial return string
1168 $sRet = '';
1169
1170 // loop through encoded strings
1171 foreach($aRet as $string) {
1172 // TODO: Do we want to control EOL (end-of-line) marker
1173 if ($sRet!='') $sRet.= " ";
1174
1175 // add header tags and encoded string to return string
1176 $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1177 }
1178
1179 return $sRet;
1180}
1181
691a2d25 1182/* This function trys to locate the entity_id of a specific mime element */
3d8371be 1183function find_ent_id($id, $message) {
a171b359 1184 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1185 if ($message->entities[$i]->header->type0 == 'multipart') {
3d8371be 1186 $ret = find_ent_id($id, $message->entities[$i]);
451f74a2 1187 } else {
3d8371be 1188 if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
d8cffbab 1189// if (sq_check_save_extension($message->entities[$i])) {
8bd0068d 1190 return $message->entities[$i]->entity_id;
da2415c1 1191// }
c8f5f606 1192 } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1193 /**
1194 * This is part of a fix for Outlook Express 6.x generating
1195 * cid URLs without creating content-id headers
1196 * @@JA - 20050207
1197 */
1198 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1199 return $message->entities[$i]->entity_id;
1200 }
3d8371be 1201 }
a3daaaf3 1202 }
451f74a2 1203 }
3d8371be 1204 return $ret;
451f74a2 1205}
a3daaaf3 1206
e5e9381a 1207function sq_check_save_extension($message) {
1208 $filename = $message->getFilename();
1209 $ext = substr($filename, strrpos($filename,'.')+1);
1210 $save_extensions = array('jpg','jpeg','gif','png','bmp');
3d8371be 1211 return in_array($ext, $save_extensions);
e5e9381a 1212}
1213
1214
691a2d25 1215/**
8bd0068d 1216 ** HTMLFILTER ROUTINES
1217 */
451f74a2 1218
691a2d25 1219/**
0493ed11 1220 * This function checks attribute values for entity-encoded values
1221 * and returns them translated into 8-bit strings so we can run
1222 * checks on them.
8bd0068d 1223 *
0493ed11 1224 * @param $attvalue A string to run entity check against.
1225 * @return Nothing, modifies a reference value.
8bd0068d 1226 */
0493ed11 1227function sq_defang(&$attvalue){
1228 $me = 'sq_defang';
2dd879b8 1229 /**
0493ed11 1230 * Skip this if there aren't ampersands or backslashes.
8bd0068d 1231 */
0493ed11 1232 if (strpos($attvalue, '&') === false
1233 && strpos($attvalue, '\\') === false){
1234 return;
2dd879b8 1235 }
0493ed11 1236 $m = false;
2b665f28 1237 // before deent, translate the dangerous unicode characters and ... to safe values
1238 // otherwise the regular expressions do not match.
1239
1240
1241
0493ed11 1242 do {
1243 $m = false;
1244 $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1245 $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1246 $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1247 } while ($m == true);
1248 $attvalue = stripslashes($attvalue);
2dd879b8 1249}
1250
1251/**
8bd0068d 1252 * Kill any tabs, newlines, or carriage returns. Our friends the
1253 * makers of the browser with 95% market value decided that it'd
1254 * be funny to make "java[tab]script" be just as good as "javascript".
1255 *
1256 * @param attvalue The attribute value before extraneous spaces removed.
0493ed11 1257 * @return attvalue Nothing, modifies a reference value.
8bd0068d 1258 */
0493ed11 1259function sq_unspace(&$attvalue){
1260 $me = 'sq_unspace';
1261 if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1262 $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1263 Array('', '', '', '', ''), $attvalue);
2dd879b8 1264 }
2dd879b8 1265}
1266
1267/**
88de4926 1268 * Translate all dangerous Unicode or Shift_JIS characters which are accepted by
2b665f28 1269 * IE as regular characters.
1270 *
1271 * @param attvalue The attribute value before dangerous characters are translated.
1272 * @return attvalue Nothing, modifies a reference value.
1273 * @author Marc Groot Koerkamp.
1274 */
1275function sq_fixIE_idiocy(&$attvalue) {
1276 // remove NUL
1277 $attvalue = str_replace("\0", "", $attvalue);
1278 // remove comments
1279 $attvalue = preg_replace("/(\/\*.*?\*\/)/","",$attvalue);
1280
88de4926 1281 // IE has the evil habit of accepting every possible value for the attribute expression.
1282 // The table below contains characters which are parsed by IE if they are used in the "expression"
2b665f28 1283 // attribute value.
1284 $aDangerousCharsReplacementTable = array(
1285 array('&#x029F;', '&#0671;' ,/* L UNICODE IPA Extension */
1286 '&#x0280;', '&#0640;' ,/* R UNICODE IPA Extension */
1287 '&#x0274;', '&#0628;' ,/* N UNICODE IPA Extension */
567dc524 1288 '&#xFF25;', '&#65317;' ,/* Unicode FULLWIDTH LATIN CAPITAL LETTER E */
1289 '&#xFF45;', '&#65349;' ,/* Unicode FULLWIDTH LATIN SMALL LETTER E */
2b665f28 1290 '&#xFF38;', '&#65336;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER X */
1291 '&#xFF58;', '&#65368;',/* Unicode FULLWIDTH LATIN SMALL LETTER X */
1292 '&#xFF30;', '&#65328;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER P */
1293 '&#xFF50;', '&#65360;',/* Unicode FULLWIDTH LATIN SMALL LETTER P */
1294 '&#xFF32;', '&#65330;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER R */
1295 '&#xFF52;', '&#65362;',/* Unicode FULLWIDTH LATIN SMALL LETTER R */
1296 '&#xFF33;', '&#65331;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER S */
1297 '&#xFF53;', '&#65363;',/* Unicode FULLWIDTH LATIN SMALL LETTER S */
1298 '&#xFF29;', '&#65321;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER I */
1299 '&#xFF49;', '&#65353;',/* Unicode FULLWIDTH LATIN SMALL LETTER I */
1300 '&#xFF2F;', '&#65327;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER O */
1301 '&#xFF4F;', '&#65359;',/* Unicode FULLWIDTH LATIN SMALL LETTER O */
1302 '&#xFF2E;', '&#65326;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER N */
1303 '&#xFF4E;', '&#65358;',/* Unicode FULLWIDTH LATIN SMALL LETTER N */
1304 '&#xFF2C;', '&#65324;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER L */
1305 '&#xFF4C;', '&#65356;',/* Unicode FULLWIDTH LATIN SMALL LETTER L */
1306 '&#xFF35;', '&#65333;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER U */
1307 '&#xFF55;', '&#65365;',/* Unicode FULLWIDTH LATIN SMALL LETTER U */
1308 '&#x207F;', '&#8319;' ,/* Unicode SUPERSCRIPT LATIN SMALL LETTER N */
567dc524 1309 "\xEF\xBC\xA5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER E */ // in unicode this is some Chinese char range
1310 "\xEF\xBD\x85", /* Shift JIS FULLWIDTH LATIN SMALL LETTER E */
1311 "\xEF\xBC\xB8", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER X */
1312 "\xEF\xBD\x98", /* Shift JIS FULLWIDTH LATIN SMALL LETTER X */
1313 "\xEF\xBC\xB0", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER P */
1314 "\xEF\xBD\x90", /* Shift JIS FULLWIDTH LATIN SMALL LETTER P */
1315 "\xEF\xBC\xB2", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER R */
1316 "\xEF\xBD\x92", /* Shift JIS FULLWIDTH LATIN SMALL LETTER R */
1317 "\xEF\xBC\xB3", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER S */
1318 "\xEF\xBD\x93", /* Shift JIS FULLWIDTH LATIN SMALL LETTER S */
1319 "\xEF\xBC\xA9", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER I */
1320 "\xEF\xBD\x89", /* Shift JIS FULLWIDTH LATIN SMALL LETTER I */
1321 "\xEF\xBC\xAF", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER O */
1322 "\xEF\xBD\x8F", /* Shift JIS FULLWIDTH LATIN SMALL LETTER O */
1323 "\xEF\xBC\xAE", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER N */
1324 "\xEF\xBD\x8E", /* Shift JIS FULLWIDTH LATIN SMALL LETTER N */
1325 "\xEF\xBC\xAC", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER L */
1326 "\xEF\xBD\x8C", /* Shift JIS FULLWIDTH LATIN SMALL LETTER L */
1327 "\xEF\xBC\xB5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER U */
1328 "\xEF\xBD\x95", /* Shift JIS FULLWIDTH LATIN SMALL LETTER U */
1329 "\xE2\x81\xBF", /* Shift JIS FULLWIDTH SUPERSCRIPT N */
1330 "\xCA\x9F", /* L UNICODE IPA Extension */
1331 "\xCA\x80", /* R UNICODE IPA Extension */
1332 "\xC9\xB4"), /* N UNICODE IPA Extension */
2b665f28 1333 array('l', 'l', 'r','r','n','n',
567dc524 1334 'E','E','e','e','X','X','x','x','P','P','p','p','R','R','r','r','S','S','s','s','I','I',
1335 'i','i','O','O','o','o','N','N','n','n','L','L','l','l','U','U','u','u','n','n',
1336 'E','e','X','x','P','p','R','r','S','s','I','i','O','o','N','n','L','l','U','u','n','l','r','n'));
2b665f28 1337 $attvalue = str_replace($aDangerousCharsReplacementTable[0],$aDangerousCharsReplacementTable[1],$attvalue);
1338
88de4926 1339 // Escapes are useful for special characters like "{}[]()'&. In other cases they are
1340 // used for XSS.
2b665f28 1341 $attvalue = preg_replace("/(\\\\)([a-zA-Z]{1})/",'$2',$attvalue);
1342}
1343
1344/**
8bd0068d 1345 * This function returns the final tag out of the tag name, an array
1346 * of attributes, and the type of the tag. This function is called by
1347 * sq_sanitize internally.
1348 *
1349 * @param $tagname the name of the tag.
1350 * @param $attary the array of attributes and their values
1351 * @param $tagtype The type of the tag (see in comments).
1352 * @return a string with the final tag representation.
1353 */
691a2d25 1354function sq_tagprint($tagname, $attary, $tagtype){
b583c3e8 1355 $me = 'sq_tagprint';
3d8371be 1356
691a2d25 1357 if ($tagtype == 2){
1358 $fulltag = '</' . $tagname . '>';
1359 } else {
1360 $fulltag = '<' . $tagname;
1361 if (is_array($attary) && sizeof($attary)){
1362 $atts = Array();
1363 while (list($attname, $attvalue) = each($attary)){
1364 array_push($atts, "$attname=$attvalue");
1365 }
1366 $fulltag .= ' ' . join(" ", $atts);
1367 }
1368 if ($tagtype == 3){
b583c3e8 1369 $fulltag .= ' /';
691a2d25 1370 }
b583c3e8 1371 $fulltag .= '>';
451f74a2 1372 }
691a2d25 1373 return $fulltag;
451f74a2 1374}
a3daaaf3 1375
691a2d25 1376/**
8bd0068d 1377 * A small helper function to use with array_walk. Modifies a by-ref
1378 * value and makes it lowercase.
1379 *
1380 * @param $val a value passed by-ref.
1381 * @return void since it modifies a by-ref value.
1382 */
691a2d25 1383function sq_casenormalize(&$val){
1384 $val = strtolower($val);
1385}
451f74a2 1386
691a2d25 1387/**
8bd0068d 1388 * This function skips any whitespace from the current position within
1389 * a string and to the next non-whitespace value.
1390 *
1391 * @param $body the string
1392 * @param $offset the offset within the string where we should start
1393 * looking for the next non-whitespace character.
1394 * @return the location within the $body where the next
1395 * non-whitespace char is located.
1396 */
691a2d25 1397function sq_skipspace($body, $offset){
b583c3e8 1398 $me = 'sq_skipspace';
3d8371be 1399 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
aab65dbd 1400 if (!empty($matches[1])){
1401 $offset += strlen($matches[1]);
451f74a2 1402 }
691a2d25 1403 return $offset;
451f74a2 1404}
a3daaaf3 1405
691a2d25 1406/**
8bd0068d 1407 * This function looks for the next character within a string. It's
1408 * really just a glorified "strpos", except it catches if failures
1409 * nicely.
1410 *
1411 * @param $body The string to look for needle in.
1412 * @param $offset Start looking from this position.
1413 * @param $needle The character/string to look for.
1414 * @return location of the next occurance of the needle, or
1415 * strlen($body) if needle wasn't found.
1416 */
691a2d25 1417function sq_findnxstr($body, $offset, $needle){
3d8371be 1418 $me = 'sq_findnxstr';
691a2d25 1419 $pos = strpos($body, $needle, $offset);
1420 if ($pos === FALSE){
1421 $pos = strlen($body);
451f74a2 1422 }
691a2d25 1423 return $pos;
451f74a2 1424}
a3daaaf3 1425
691a2d25 1426/**
8bd0068d 1427 * This function takes a PCRE-style regexp and tries to match it
1428 * within the string.
1429 *
1430 * @param $body The string to look for needle in.
1431 * @param $offset Start looking from here.
1432 * @param $reg A PCRE-style regex to match.
1433 * @return Returns a false if no matches found, or an array
1434 * with the following members:
1435 * - integer with the location of the match within $body
1436 * - string with whatever content between offset and the match
1437 * - string with whatever it is we matched
1438 */
691a2d25 1439function sq_findnxreg($body, $offset, $reg){
b583c3e8 1440 $me = 'sq_findnxreg';
691a2d25 1441 $matches = Array();
1442 $retarr = Array();
7d06541f 1443 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1444 if (!isset($matches{0}) || !$matches{0}){
691a2d25 1445 $retarr = false;
1446 } else {
1447 $retarr{0} = $offset + strlen($matches{1});
1448 $retarr{1} = $matches{1};
1449 $retarr{2} = $matches{2};
1450 }
1451 return $retarr;
1452}
a3daaaf3 1453
691a2d25 1454/**
8bd0068d 1455 * This function looks for the next tag.
1456 *
1457 * @param $body String where to look for the next tag.
1458 * @param $offset Start looking from here.
1459 * @return false if no more tags exist in the body, or
1460 * an array with the following members:
1461 * - string with the name of the tag
1462 * - array with attributes and their values
1463 * - integer with tag type (1, 2, or 3)
1464 * - integer where the tag starts (starting "<")
1465 * - integer where the tag ends (ending ">")
1466 * first three members will be false, if the tag is invalid.
1467 */
691a2d25 1468function sq_getnxtag($body, $offset){
b583c3e8 1469 $me = 'sq_getnxtag';
691a2d25 1470 if ($offset > strlen($body)){
1471 return false;
1472 }
1473 $lt = sq_findnxstr($body, $offset, "<");
1474 if ($lt == strlen($body)){
1475 return false;
1476 }
1477 /**
8bd0068d 1478 * We are here:
1479 * blah blah <tag attribute="value">
1480 * \---------^
1481 */
691a2d25 1482 $pos = sq_skipspace($body, $lt+1);
1483 if ($pos >= strlen($body)){
1484 return Array(false, false, false, $lt, strlen($body));
1485 }
1486 /**
8bd0068d 1487 * There are 3 kinds of tags:
1488 * 1. Opening tag, e.g.:
1489 * <a href="blah">
1490 * 2. Closing tag, e.g.:
1491 * </a>
1492 * 3. XHTML-style content-less tag, e.g.:
1493 * <img src="blah" />
1494 */
691a2d25 1495 $tagtype = false;
1496 switch (substr($body, $pos, 1)){
3d8371be 1497 case '/':
1498 $tagtype = 2;
1499 $pos++;
1500 break;
1501 case '!':
1502 /**
8bd0068d 1503 * A comment or an SGML declaration.
1504 */
3d8371be 1505 if (substr($body, $pos+1, 2) == "--"){
1506 $gt = strpos($body, "-->", $pos);
1507 if ($gt === false){
1508 $gt = strlen($body);
1509 } else {
1510 $gt += 2;
1511 }
1512 return Array(false, false, false, $lt, $gt);
bb8d0799 1513 } else {
3d8371be 1514 $gt = sq_findnxstr($body, $pos, ">");
1515 return Array(false, false, false, $lt, $gt);
1516 }
1517 break;
1518 default:
1519 /**
8bd0068d 1520 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1521 * later.
1522 */
3d8371be 1523 $tagtype = 1;
1524 break;
691a2d25 1525 }
a3daaaf3 1526
0493ed11 1527 $tag_start = $pos;
691a2d25 1528 $tagname = '';
1529 /**
8bd0068d 1530 * Look for next [\W-_], which will indicate the end of the tag name.
1531 */
691a2d25 1532 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1533 if ($regary == false){
1534 return Array(false, false, false, $lt, strlen($body));
1535 }
1536 list($pos, $tagname, $match) = $regary;
1537 $tagname = strtolower($tagname);
1538
1539 /**
8bd0068d 1540 * $match can be either of these:
1541 * '>' indicating the end of the tag entirely.
1542 * '\s' indicating the end of the tag name.
1543 * '/' indicating that this is type-3 xhtml tag.
1544 *
1545 * Whatever else we find there indicates an invalid tag.
1546 */
691a2d25 1547 switch ($match){
3d8371be 1548 case '/':
691a2d25 1549 /**
8bd0068d 1550 * This is an xhtml-style tag with a closing / at the
1551 * end, like so: <img src="blah" />. Check if it's followed
1552 * by the closing bracket. If not, then this tag is invalid
1553 */
3d8371be 1554 if (substr($body, $pos, 2) == "/>"){
1555 $pos++;
1556 $tagtype = 3;
1557 } else {
1558 $gt = sq_findnxstr($body, $pos, ">");
1559 $retary = Array(false, false, false, $lt, $gt);
1560 return $retary;
1561 }
1562 case '>':
1563 return Array($tagname, false, $tagtype, $lt, $pos);
1564 break;
1565 default:
1566 /**
8bd0068d 1567 * Check if it's whitespace
1568 */
3d8371be 1569 if (!preg_match('/\s/', $match)){
1570 /**
8bd0068d 1571 * This is an invalid tag! Look for the next closing ">".
1572 */
7d06541f 1573 $gt = sq_findnxstr($body, $lt, ">");
3d8371be 1574 return Array(false, false, false, $lt, $gt);
1575 }
1576 break;
691a2d25 1577 }
3d8371be 1578
691a2d25 1579 /**
8bd0068d 1580 * At this point we're here:
1581 * <tagname attribute='blah'>
1582 * \-------^
1583 *
1584 * At this point we loop in order to find all attributes.
1585 */
691a2d25 1586 $attname = '';
0493ed11 1587 $atttype = false;
691a2d25 1588 $attary = Array();
1589
1590 while ($pos <= strlen($body)){
1591 $pos = sq_skipspace($body, $pos);
1592 if ($pos == strlen($body)){
1593 /**
8bd0068d 1594 * Non-closed tag.
1595 */
691a2d25 1596 return Array(false, false, false, $lt, $pos);
1597 }
1598 /**
8bd0068d 1599 * See if we arrived at a ">" or "/>", which means that we reached
1600 * the end of the tag.
1601 */
691a2d25 1602 $matches = Array();
164800ad 1603 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
c828931c 1604 /**
8bd0068d 1605 * Yep. So we did.
1606 */
c828931c 1607 $pos += strlen($matches{1});
1608 if ($matches{2} == "/>"){
1609 $tagtype = 3;
1610 $pos++;
1611 }
1612 return Array($tagname, $attary, $tagtype, $lt, $pos);
1613 }
a3daaaf3 1614
cca46357 1615 /**
8bd0068d 1616 * There are several types of attributes, with optional
1617 * [:space:] between members.
1618 * Type 1:
1619 * attrname[:space:]=[:space:]'CDATA'
1620 * Type 2:
1621 * attrname[:space:]=[:space:]"CDATA"
1622 * Type 3:
1623 * attr[:space:]=[:space:]CDATA
1624 * Type 4:
1625 * attrname
1626 *
1627 * We leave types 1 and 2 the same, type 3 we check for
1628 * '"' and convert to "&quot" if needed, then wrap in
1629 * double quotes. Type 4 we convert into:
1630 * attrname="yes".
1631 */
3f7c623f 1632 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
691a2d25 1633 if ($regary == false){
1634 /**
8bd0068d 1635 * Looks like body ended before the end of tag.
1636 */
691a2d25 1637 return Array(false, false, false, $lt, strlen($body));
cca46357 1638 }
691a2d25 1639 list($pos, $attname, $match) = $regary;
1640 $attname = strtolower($attname);
1641 /**
8bd0068d 1642 * We arrived at the end of attribute name. Several things possible
1643 * here:
1644 * '>' means the end of the tag and this is attribute type 4
1645 * '/' if followed by '>' means the same thing as above
1646 * '\s' means a lot of things -- look what it's followed by.
1647 * anything else means the attribute is invalid.
1648 */
691a2d25 1649 switch($match){
3d8371be 1650 case '/':
691a2d25 1651 /**
8bd0068d 1652 * This is an xhtml-style tag with a closing / at the
1653 * end, like so: <img src="blah" />. Check if it's followed
1654 * by the closing bracket. If not, then this tag is invalid
1655 */
3d8371be 1656 if (substr($body, $pos, 2) == "/>"){
691a2d25 1657 $pos++;
3d8371be 1658 $tagtype = 3;
691a2d25 1659 } else {
3d8371be 1660 $gt = sq_findnxstr($body, $pos, ">");
1661 $retary = Array(false, false, false, $lt, $gt);
1662 return $retary;
1663 }
1664 case '>':
1665 $attary{$attname} = '"yes"';
1666 return Array($tagname, $attary, $tagtype, $lt, $pos);
1667 break;
1668 default:
1669 /**
8bd0068d 1670 * Skip whitespace and see what we arrive at.
1671 */
3d8371be 1672 $pos = sq_skipspace($body, $pos);
1673 $char = substr($body, $pos, 1);
1674 /**
8bd0068d 1675 * Two things are valid here:
1676 * '=' means this is attribute type 1 2 or 3.
1677 * \w means this was attribute type 4.
1678 * anything else we ignore and re-loop. End of tag and
1679 * invalid stuff will be caught by our checks at the beginning
1680 * of the loop.
1681 */
3d8371be 1682 if ($char == "="){
1683 $pos++;
1684 $pos = sq_skipspace($body, $pos);
691a2d25 1685 /**
8bd0068d 1686 * Here are 3 possibilities:
1687 * "'" attribute type 1
1688 * '"' attribute type 2
1689 * everything else is the content of tag type 3
1690 */
3d8371be 1691 $quot = substr($body, $pos, 1);
1692 if ($quot == "'"){
1693 $regary = sq_findnxreg($body, $pos+1, "\'");
1694 if ($regary == false){
1695 return Array(false, false, false, $lt, strlen($body));
1696 }
1697 list($pos, $attval, $match) = $regary;
1698 $pos++;
1699 $attary{$attname} = "'" . $attval . "'";
1700 } else if ($quot == '"'){
1701 $regary = sq_findnxreg($body, $pos+1, '\"');
1702 if ($regary == false){
1703 return Array(false, false, false, $lt, strlen($body));
1704 }
1705 list($pos, $attval, $match) = $regary;
1706 $pos++;
1707 $attary{$attname} = '"' . $attval . '"';
1708 } else {
1709 /**
8bd0068d 1710 * These are hateful. Look for \s, or >.
1711 */
3d8371be 1712 $regary = sq_findnxreg($body, $pos, "[\s>]");
1713 if ($regary == false){
1714 return Array(false, false, false, $lt, strlen($body));
1715 }
1716 list($pos, $attval, $match) = $regary;
1717 /**
8bd0068d 1718 * If it's ">" it will be caught at the top.
1719 */
3d8371be 1720 $attval = preg_replace("/\"/s", "&quot;", $attval);
1721 $attary{$attname} = '"' . $attval . '"';
7e235a1a 1722 }
3d8371be 1723 } else if (preg_match("|[\w/>]|", $char)) {
691a2d25 1724 /**
8bd0068d 1725 * That was attribute type 4.
1726 */
3d8371be 1727 $attary{$attname} = '"yes"';
1728 } else {
1729 /**
8bd0068d 1730 * An illegal character. Find next '>' and return.
1731 */
3d8371be 1732 $gt = sq_findnxstr($body, $pos, ">");
1733 return Array(false, false, false, $lt, $gt);
451f74a2 1734 }
3d8371be 1735 break;
691a2d25 1736 }
1737 }
1738 /**
8bd0068d 1739 * The fact that we got here indicates that the tag end was never
1740 * found. Return invalid tag indication so it gets stripped.
1741 */
691a2d25 1742 return Array(false, false, false, $lt, strlen($body));
1743}
1744
1745/**
0493ed11 1746 * Translates entities into literal values so they can be checked.
8bd0068d 1747 *
0493ed11 1748 * @param $attvalue the by-ref value to check.
1749 * @param $regex the regular expression to check against.
1750 * @param $hex whether the entites are hexadecimal.
1751 * @return True or False depending on whether there were matches.
8bd0068d 1752 */
0493ed11 1753function sq_deent(&$attvalue, $regex, $hex=false){
b583c3e8 1754 $me = 'sq_deent';
0493ed11 1755 $ret_match = false;
2b665f28 1756 // remove comments
1757 //$attvalue = preg_replace("/(\/\*.*\*\/)/","",$attvalue);
0493ed11 1758 preg_match_all($regex, $attvalue, $matches);
1759 if (is_array($matches) && sizeof($matches[0]) > 0){
1760 $repl = Array();
1761 for ($i = 0; $i < sizeof($matches[0]); $i++){
1762 $numval = $matches[1][$i];
1763 if ($hex){
1764 $numval = hexdec($numval);
a3daaaf3 1765 }
0493ed11 1766 $repl{$matches[0][$i]} = chr($numval);
691a2d25 1767 }
0493ed11 1768 $attvalue = strtr($attvalue, $repl);
1769 return true;
1770 } else {
1771 return false;
691a2d25 1772 }
691a2d25 1773}
1774
1775/**
8bd0068d 1776 * This function runs various checks against the attributes.
1777 *
1778 * @param $tagname String with the name of the tag.
1779 * @param $attary Array with all tag attributes.
1780 * @param $rm_attnames See description for sq_sanitize
1781 * @param $bad_attvals See description for sq_sanitize
1782 * @param $add_attr_to_tag See description for sq_sanitize
1783 * @param $message message object
1784 * @param $id message id
1785 * @return Array with modified attributes.
1786 */
da2415c1 1787function sq_fixatts($tagname,
1788 $attary,
691a2d25 1789 $rm_attnames,
1790 $bad_attvals,
1791 $add_attr_to_tag,
1792 $message,
b3af12ef 1793 $id,
3d8371be 1794 $mailbox
691a2d25 1795 ){
b583c3e8 1796 $me = 'sq_fixatts';
691a2d25 1797 while (list($attname, $attvalue) = each($attary)){
1798 /**
8bd0068d 1799 * See if this attribute should be removed.
1800 */
691a2d25 1801 foreach ($rm_attnames as $matchtag=>$matchattrs){
1802 if (preg_match($matchtag, $tagname)){
1803 foreach ($matchattrs as $matchattr){
1804 if (preg_match($matchattr, $attname)){
1805 unset($attary{$attname});
1806 continue;
1807 }
451f74a2 1808 }
451f74a2 1809 }
691a2d25 1810 }
1811 /**
2b665f28 1812 * Workaround for IE quirks
1813 */
1814 sq_fixIE_idiocy($attvalue);
1815
1816 /**
8bd0068d 1817 * Remove any backslashes, entities, and extraneous whitespace.
1818 */
2b665f28 1819
1820 $oldattvalue = $attvalue;
0493ed11 1821 sq_defang($attvalue);
2b665f28 1822 if ($attname == 'style' && $attvalue !== $oldattvalue) {
1823 // entities are used in the attribute value. In 99% of the cases it's there as XSS
1824 // i.e.<div style="{ left:exp&#x0280;essio&#x0274;( alert('XSS') ) }">
1825 $attvalue = "idiocy";
1826 $attary{$attname} = $attvalue;
1827 }
0493ed11 1828 sq_unspace($attvalue);
af861a34 1829
1830 /**
8bd0068d 1831 * Now let's run checks on the attvalues.
1832 * I don't expect anyone to comprehend this. If you do,
1833 * get in touch with me so I can drive to where you live and
1834 * shake your hand personally. :)
1835 */
691a2d25 1836 foreach ($bad_attvals as $matchtag=>$matchattrs){
1837 if (preg_match($matchtag, $tagname)){
1838 foreach ($matchattrs as $matchattr=>$valary){
1839 if (preg_match($matchattr, $attname)){
1840 /**
8bd0068d 1841 * There are two arrays in valary.
1842 * First is matches.
1843 * Second one is replacements
1844 */
691a2d25 1845 list($valmatch, $valrepl) = $valary;
da2415c1 1846 $newvalue =
691a2d25 1847 preg_replace($valmatch, $valrepl, $attvalue);
1848 if ($newvalue != $attvalue){
1849 $attary{$attname} = $newvalue;
567dc524 1850 $attvalue = $newvalue;
691a2d25 1851 }
1852 }
1853 }
451f74a2 1854 }
a3daaaf3 1855 }
567dc524 1856 if ($attname == 'style') {
1857 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1858 // 8bit and control characters in style attribute values can be used for XSS, remove them
1859 $attary{$attname} = '"disallowed character"';
1860 }
1861 preg_match_all("/url\s*\((.+)\)/si",$attvalue,$aMatch);
1862 if (count($aMatch)) {
1863 foreach($aMatch[1] as $sMatch) {
1864 // url value
1865 $urlvalue = $sMatch;
1866 sq_fix_url($attname, $urlvalue, $message, $id, $mailbox,"'");
1867 $attary{$attname} = str_replace($sMatch,$urlvalue,$attvalue);
1868 }
1869 }
691a2d25 1870 }
ff940ebc 1871 /**
567dc524 1872 * Use white list based filtering on attributes which can contain url's
ff940ebc 1873 */
caf0ab1d 1874 else if ($attname == 'href' || $attname == 'xlink:href' || $attname == 'src'
1875 || $attname == 'poster' || $attname == 'formaction'
1876 || $attname == 'background' || $attname == 'action') {
567dc524 1877 sq_fix_url($attname, $attvalue, $message, $id, $mailbox);
1878 $attary{$attname} = $attvalue;
ff940ebc 1879 }
a3daaaf3 1880 }
691a2d25 1881 /**
8bd0068d 1882 * See if we need to append any attributes to this tag.
1883 */
691a2d25 1884 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1885 if (preg_match($matchtag, $tagname)){
1886 $attary = array_merge($attary, $addattary);
1887 }
1888 }
1889 return $attary;
451f74a2 1890}
a3daaaf3 1891
691a2d25 1892/**
567dc524 1893 * This function filters url's
1894 *
1895 * @param $attvalue String with attribute value to filter
1896 * @param $message message object
1897 * @param $id message id
1898 * @param $mailbox mailbox
1899 * @param $sQuote quoting characters around url's
1900 */
1901function sq_fix_url($attname, &$attvalue, $message, $id, $mailbox,$sQuote = '"') {
1902 $attvalue = trim($attvalue);
1903 if ($attvalue && ($attvalue[0] =='"'|| $attvalue[0] == "'")) {
1904 // remove the double quotes
1905 $sQuote = $attvalue[0];
1906 $attvalue = trim(substr($attvalue,1,-1));
1907 }
1908
d31f73f1 1909 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
1910 // images off by default.
afbf184c 1911 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET, FALSE);
d31f73f1 1912
6ab8321f 1913 global $use_transparent_security_image;
1914 if ($use_transparent_security_image) $secremoveimg = '../images/spacer.png';
1915 else $secremoveimg = '../images/' . _("sec_remove_eng.png");
567dc524 1916
1917 /**
1918 * Replace empty src tags with the blank image. src is only used
1919 * for frames, images, and image inputs. Doing a replace should
1920 * not affect them working as should be, however it will stop
1921 * IE from being kicked off when src for img tags are not set
1922 */
1923 if ($attvalue == '') {
1924 $attvalue = '"' . SM_PATH . 'images/blank.png"';
1925 } else {
1926 // first, disallow 8 bit characters and control characters
1927 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1928 switch ($attname) {
1929 case 'href':
1930 $attvalue = $sQuote . 'http://invalid-stuff-detected.example.com' . $sQuote;
1931 break;
1932 default:
1933 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1934 break;
1935 }
1936 } else {
1937 $aUrl = parse_url($attvalue);
1938 if (isset($aUrl['scheme'])) {
1939 switch(strtolower($aUrl['scheme'])) {
d75e755b 1940 case 'mailto':
567dc524 1941 case 'http':
1942 case 'https':
1943 case 'ftp':
1944 if ($attname != 'href') {
1945 if ($view_unsafe_images == false) {
1946 $attvalue = $sQuote . $secremoveimg . $sQuote;
1947 } else {
1948 if (isset($aUrl['path'])) {
8bbbc757 1949
1950 // No one has been able to show that image URIs
1951 // can be exploited, so for now, no restrictions
1952 // are made at all. If this proves to be a problem,
1953 // the commented-out code below can be of help.
1954 // (One consideration is that I see nothing in this
1955 // function that specifically says that we will
1956 // only ever arrive here when inspecting an image
1957 // tag, although that does seem to be the end
1958 // result - e.g., <script src="..."> where malicious
1959 // image URIs are in fact a problem are already
1960 // filtered out elsewhere.
1961 /* ---------------------------------
567dc524 1962 // validate image extension.
1963 $ext = strtolower(substr($aUrl['path'],strrpos($aUrl['path'],'.')));
1964 if (!in_array($ext,array('.jpeg','.jpg','xjpeg','.gif','.bmp','.jpe','.png','.xbm'))) {
8bbbc757 1965 // If URI is to something other than
1966 // a regular image file, get the contents
1967 // and try to see if it is an image.
1968 // Don't use Fileinfo (finfo_file()) because
1969 // we'd need to make the admin configure the
1970 // location of the magic.mime file (FIXME: add finfo_file() support later?)
1971 //
1972 $mime_type = '';
1973 if (function_exists('mime_content_type')
1974 && ($FILE = @fopen($attvalue, 'rb', FALSE))) {
1975
1976 // fetch file
1977 //
1978 $file_contents = '';
1979 while (!feof($FILE)) {
1980 $file_contents .= fread($FILE, 8192);
1981 }
1982 fclose($FILE);
1983
1984 // store file locally
1985 //
1986 global $attachment_dir, $username;
1987 $hashed_attachment_dir = getHashedDir($username, $attachment_dir);
1988 $localfilename = GenerateRandomString(32, '', 7);
1989 $full_localfilename = "$hashed_attachment_dir/$localfilename";
1990 while (file_exists($full_localfilename)) {
1991 $localfilename = GenerateRandomString(32, '', 7);
1992 $full_localfilename = "$hashed_attachment_dir/$localfilename";
1993 }
1994 $FILE = fopen("$hashed_attachment_dir/$localfilename", 'wb');
1995 fwrite($FILE, $file_contents);
1996 fclose($FILE);
1997
1998 // get mime type and remove file
1999 //
2000 $mime_type = mime_content_type("$hashed_attachment_dir/$localfilename");
2001 unlink("$hashed_attachment_dir/$localfilename");
2002 }
2003 // debug: echo "$attvalue FILE TYPE IS $mime_type<HR>";
2004 if (substr(strtolower($mime_type), 0, 5) != 'image') {
2005 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
2006 }
567dc524 2007 }
8bbbc757 2008 --------------------------------- */
567dc524 2009 } else {
2010 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
2011 }
2012 }
cf088f55 2013 } else {
2014 $attvalue = $sQuote . $attvalue . $sQuote;
567dc524 2015 }
2016 break;
2017 case 'outbind':
2018 /**
2019 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
2020 * One day MS might actually make it match something useful, for now, falling
2021 * back to using cid2http, so we can grab the blank.png.
2022 */
cf088f55 2023 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
567dc524 2024 break;
2025 case 'cid':
2026 /**
2027 * Turn cid: urls into http-friendly ones.
2028 */
cf088f55 2029 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
567dc524 2030 break;
2031 default:
2032 $attvalue = $sQuote . SM_PATH . 'images/blank.png' . $sQuote;
2033 break;
2034 }
2035 } else {
b1fbb25f 2036 if (!isset($aUrl['path']) || $aUrl['path'] != $secremoveimg) {
567dc524 2037 // parse_url did not lead to satisfying result
2038 $attvalue = $sQuote . SM_PATH . 'images/blank.png' . $sQuote;
2039 }
2040 }
2041 }
2042 }
2043}
2044
2045/**
8bd0068d 2046 * This function edits the style definition to make them friendly and
2047 * usable in SquirrelMail.
2048 *
2049 * @param $message the message object
2050 * @param $id the message id
2051 * @param $content a string with whatever is between <style> and </style>
2052 * @param $mailbox the message mailbox
2053 * @return a string with edited content.
2054 */
e60a299a 2055function sq_fixstyle($body, $pos, $message, $id, $mailbox){
b583c3e8 2056 $me = 'sq_fixstyle';
7e2ff844 2057 // workaround for </style> in between comments
2058 $iCurrentPos = $pos;
2059 $content = '';
2060 $sToken = '';
2061 $bSucces = false;
2062 $bEndTag = false;
2063 for ($i=$pos,$iCount=strlen($body);$i<$iCount;++$i) {
2064 $char = $body{$i};
2065 switch ($char) {
2066 case '<':
10c0caeb 2067 $sToken = $char;
7e2ff844 2068 break;
2069 case '/':
2070 if ($sToken == '<') {
2071 $sToken .= $char;
2072 $bEndTag = true;
2073 } else {
2074 $content .= $char;
2075 }
2076 break;
2077 case '>':
2078 if ($bEndTag) {
2079 $sToken .= $char;
2080 if (preg_match('/\<\/\s*style\s*\>/i',$sToken,$aMatch)) {
2081 $newpos = $i + 1;
2082 $bSucces = true;
2083 break 2;
2084 } else {
2085 $content .= $sToken;
2086 }
2087 $bEndTag = false;
2088 } else {
2089 $content .= $char;
2090 }
2091 break;
2092 case '!':
2093 if ($sToken == '<') {
2094 // possible comment
2095 if (isset($body{$i+2}) && substr($body,$i,3) == '!--') {
2096 $i = strpos($body,'-->',$i+3);
2b665f28 2097 if ($i === false) { // no end comment
2098 $i = strlen($body);
2099 }
7e2ff844 2100 $sToken = '';
2101 }
2102 } else {
2103 $content .= $char;
2104 }
2105 break;
2106 default:
2107 if ($bEndTag) {
2108 $sToken .= $char;
2109 } else {
2110 $content .= $char;
2111 }
2112 break;
2113 }
2114 }
2115 if ($bSucces == FALSE){
7d06541f 2116 return array(FALSE, strlen($body));
2117 }
7e2ff844 2118
2b665f28 2119
2120
691a2d25 2121 /**
9ebace19 2122 * First look for general BODY style declaration, which would be
2123 * like so:
2124 * body {background: blah-blah}
2125 * and change it to .bodyclass so we can just assign it to a <div>
2126 */
691a2d25 2127 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
6ab8321f 2128
2129 global $use_transparent_security_image;
2130 if ($use_transparent_security_image) $secremoveimg = '../images/spacer.png';
2131 else $secremoveimg = '../images/' . _("sec_remove_eng.png");
2132
691a2d25 2133 /**
0493ed11 2134 * Fix url('blah') declarations.
2135 */
2136 // $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
2137 // "url(\\1$secremoveimg\\2)", $content);
2b665f28 2138
567dc524 2139 // first check for 8bit sequences and disallowed control characters
2140 if (preg_match('/[\16-\37\200-\377]+/',$content)) {
2141 $content = '<!-- style block removed by html filter due to presence of 8bit characters -->';
2142 return array($content, $newpos);
2143 }
2144
2b665f28 2145 // IE Sucks hard. We have a special function for it.
2146 sq_fixIE_idiocy($content);
2147
2148 // remove @import line
2149 $content = preg_replace("/^\s*(@import.*)$/mi","\n<!-- @import rules forbidden -->\n",$content);
2150
5b4884be 2151 // translate ur\l and variations (IE parses that)
2b665f28 2152 // TODO check if the sq_fixIE_idiocy function already handles this.
5b4884be 2153 $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
567dc524 2154 preg_match_all("/url\s*\((.+)\)/si",$content,$aMatch);
2155 if (count($aMatch)) {
2156 $aValue = $aReplace = array();
2157 foreach($aMatch[1] as $sMatch) {
2158 // url value
2159 $urlvalue = $sMatch;
2160 sq_fix_url('style',$urlvalue, $message, $id, $mailbox,"'");
2161 $aValue[] = $sMatch;
2162 $aReplace[] = $urlvalue;
0493ed11 2163 }
567dc524 2164 $content = str_replace($aValue,$aReplace,$content);
691a2d25 2165 }
567dc524 2166
9ebace19 2167 /**
2168 * Remove any backslashes, entities, and extraneous whitespace.
2169 */
0493ed11 2170 $contentTemp = $content;
2171 sq_defang($contentTemp);
2172 sq_unspace($contentTemp);
a3daaaf3 2173
691a2d25 2174 /**
8bd0068d 2175 * Fix stupid css declarations which lead to vulnerabilities
2176 * in IE.
39352565 2177 *
2178 * Also remove "position" attribute, as it can easily be set
2179 * to "fixed" or "absolute" with "left" and "top" attributes
2180 * of zero, taking over the whole content frame. It can also
2181 * be set to relative and move itself anywhere it wants to,
2182 * displaying content in areas it shouldn't be allowed to touch.
8bd0068d 2183 */
caf0ab1d 2184 $match = Array('/\/\*.*\*\//', // removes /* blah blah */
5db90261 2185 '/expression/i',
0493ed11 2186 '/behaviou*r/i',
2187 '/binding/i',
2b665f28 2188 '/include-source/i',
2189 '/javascript/i',
39352565 2190 '/script/i',
2191 '/position/i');
2192 $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', '');
0493ed11 2193 $contentNew = preg_replace($match, $replace, $contentTemp);
2194 if ($contentNew !== $contentTemp) {
2195 // insecure css declarations are used. From now on we don't care
2196 // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
2197 $content = $contentNew;
2198 }
7d06541f 2199 return array($content, $newpos);
691a2d25 2200}
a3daaaf3 2201
0493ed11 2202
691a2d25 2203/**
8bd0068d 2204 * This function converts cid: url's into the ones that can be viewed in
2205 * the browser.
2206 *
2207 * @param $message the message object
2208 * @param $id the message id
2209 * @param $cidurl the cid: url.
2210 * @param $mailbox the message mailbox
2211 * @return a string with a http-friendly url
2212 */
b3af12ef 2213function sq_cid2http($message, $id, $cidurl, $mailbox){
691a2d25 2214 /**
8bd0068d 2215 * Get rid of quotes.
2216 */
691a2d25 2217 $quotchar = substr($cidurl, 0, 1);
2dd879b8 2218 if ($quotchar == '"' || $quotchar == "'"){
2219 $cidurl = str_replace($quotchar, "", $cidurl);
2220 } else {
2221 $quotchar = '';
2222 }
691a2d25 2223 $cidurl = substr(trim($cidurl), 4);
0493ed11 2224
2225 $match_str = '/\{.*?\}\//';
2226 $str_rep = '';
2227 $cidurl = preg_replace($match_str, $str_rep, $cidurl);
2228
e5e9381a 2229 $linkurl = find_ent_id($cidurl, $message);
9ebace19 2230 /* in case of non-safe cid links $httpurl should be replaced by a sort of
2231 unsafe link image */
e5e9381a 2232 $httpurl = '';
c8f5f606 2233
8bd0068d 2234 /**
2235 * This is part of a fix for Outlook Express 6.x generating
2236 * cid URLs without creating content-id headers. These images are
2237 * not part of the multipart/related html mail. The html contains
2238 * <img src="cid:{some_id}/image_filename.ext"> references to
2239 * attached images with as goal to render them inline although
2240 * the attachment disposition property is not inline.
2241 */
c8f5f606 2242
2243 if (empty($linkurl)) {
2244 if (preg_match('/{.*}\//', $cidurl)) {
2245 $cidurl = preg_replace('/{.*}\//','', $cidurl);
2246 if (!empty($cidurl)) {
2247 $linkurl = find_ent_id($cidurl, $message);
2248 }
2249 }
2250 }
f8a1ed5a 2251
c8f5f606 2252 if (!empty($linkurl)) {
fa26ab45 2253 $httpurl = $quotchar . sqm_baseuri() . 'src/download.php?absolute_dl=true&amp;' .
8bd0068d 2254 "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
2255 '&amp;ent_id=' . $linkurl . $quotchar;
c8f5f606 2256 } else {
2257 /**
2258 * If we couldn't generate a proper img url, drop in a blank image
2259 * instead of sending back empty, otherwise it causes unusual behaviour
2260 */
bc017c1d 2261 $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
e5e9381a 2262 }
f8a1ed5a 2263
691a2d25 2264 return $httpurl;
2265}
2266
2267/**
8bd0068d 2268 * This function changes the <body> tag into a <div> tag since we
2269 * can't really have a body-within-body.
2270 *
2271 * @param $attary an array of attributes and values of <body>
2272 * @param $mailbox mailbox we're currently reading (for cid2http)
2273 * @param $message current message (for cid2http)
2274 * @param $id current message id (for cid2http)
2275 * @return a modified array of attributes to be set for <div>
2276 */
2dd879b8 2277function sq_body2div($attary, $mailbox, $message, $id){
b583c3e8 2278 $me = 'sq_body2div';
3d8371be 2279 $divattary = Array('class' => "'bodyclass'");
b583c3e8 2280 $text = '#000000';
c189a963 2281 $has_bgc_stl = $has_txt_stl = false;
b583c3e8 2282 $styledef = '';
691a2d25 2283 if (is_array($attary) && sizeof($attary) > 0){
2284 foreach ($attary as $attname=>$attvalue){
2285 $quotchar = substr($attvalue, 0, 1);
2286 $attvalue = str_replace($quotchar, "", $attvalue);
2287 switch ($attname){
3d8371be 2288 case 'background':
8bd0068d 2289 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
3d8371be 2290 $styledef .= "background-image: url('$attvalue'); ";
2291 break;
2292 case 'bgcolor':
c189a963 2293 $has_bgc_stl = true;
3d8371be 2294 $styledef .= "background-color: $attvalue; ";
2295 break;
2296 case 'text':
c189a963 2297 $has_txt_stl = true;
3d8371be 2298 $styledef .= "color: $attvalue; ";
2299 break;
691a2d25 2300 }
a3daaaf3 2301 }
c189a963 2302 // Outlook defines a white bgcolor and no text color. This can lead to
2303 // white text on a white bg with certain themes.
2304 if ($has_bgc_stl && !$has_txt_stl) {
2305 $styledef .= "color: $text; ";
2306 }
691a2d25 2307 if (strlen($styledef) > 0){
2308 $divattary{"style"} = "\"$styledef\"";
2309 }
2310 }
2311 return $divattary;
2312}
a3daaaf3 2313
691a2d25 2314/**
8bd0068d 2315 * This is the main function and the one you should actually be calling.
2316 * There are several variables you should be aware of an which need
2317 * special description.
2318 *
2319 * Since the description is quite lengthy, see it here:
2320 * http://linux.duke.edu/projects/mini/htmlfilter/
2321 *
2322 * @param $body the string with HTML you wish to filter
2323 * @param $tag_list see description above
2324 * @param $rm_tags_with_content see description above
2325 * @param $self_closing_tags see description above
2326 * @param $force_tag_closing see description above
2327 * @param $rm_attnames see description above
2328 * @param $bad_attvals see description above
2329 * @param $add_attr_to_tag see description above
2330 * @param $message message object
2331 * @param $id message id
dfc64d14 2332 * @param $recursively_called boolean flag for recursive calls into this function (optional; default FALSE)
8bd0068d 2333 * @return sanitized html safe to show on your pages.
2334 */
da2415c1 2335function sq_sanitize($body,
8bd0068d 2336 $tag_list,
2337 $rm_tags_with_content,
2338 $self_closing_tags,
2339 $force_tag_closing,
2340 $rm_attnames,
2341 $bad_attvals,
2342 $add_attr_to_tag,
2343 $message,
2344 $id,
dfc64d14 2345 $mailbox,
2346 $recursively_called=FALSE
8bd0068d 2347 ){
b583c3e8 2348 $me = 'sq_sanitize';
dfc64d14 2349
2350 /**
2351 * See if tag_list is of tags to remove or tags to allow.
2352 * false means remove these tags
2353 * true means allow these tags
2354 */
2355 $orig_tag_list = $tag_list;
7d06541f 2356 $rm_tags = array_shift($tag_list);
dfc64d14 2357
691a2d25 2358 /**
8bd0068d 2359 * Normalize rm_tags and rm_tags_with_content.
2360 */
7d06541f 2361 @array_walk($tag_list, 'sq_casenormalize');
691a2d25 2362 @array_walk($rm_tags_with_content, 'sq_casenormalize');
2363 @array_walk($self_closing_tags, 'sq_casenormalize');
dfc64d14 2364
691a2d25 2365 $curpos = 0;
2366 $open_tags = Array();
2dd879b8 2367 $trusted = "\n<!-- begin sanitized html -->\n";
691a2d25 2368 $skip_content = false;
bb8d0799 2369 /**
8bd0068d 2370 * Take care of netscape's stupid javascript entities like
2371 * &{alert('boo')};
2372 */
bb8d0799 2373 $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
691a2d25 2374
7d06541f 2375 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
691a2d25 2376 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
dfc64d14 2377
2378 /**
2379 * RCDATA and RAWTEXT tags are handled differently:
2380 * next instance of closing tag is used, whether or not
2381 * the HTML is well formed before that
2382 */
2383 global $rcdata_rawtext_tags;
2384 if (!$recursively_called
2385 && in_array($tagname, $rcdata_rawtext_tags)
2386 && $tagtype === 1){
2387 $closing_tag = false;
2388 $closing_tag_offset = $curpos;
2389 // seek out the closing tag for the current RCDATA/RAWTEXT tag
2390 while (1) {
2391 // first we need to move forward to next available closing tag
2392 // (intentionally leave off the closing > and let sq_getnxtag() validate a proper tag syntax)
2393 $next_tag = sq_findnxreg($body, $closing_tag_offset, "</\s*$tagname");
2394 if ($next_tag === false) {
2395 $closing_tag = false;
2396 break;
2397 }
2398 // but then we have to make sure it's a well-formed tag
2399 $closing_tag = sq_getnxtag($body, $next_tag[0]);
2400 if ($closing_tag === false)
2401 break;
2402 else if ($closing_tag[0] !== false
2403 // these should be redundant
2404 && $closing_tag[0] === $tagname && $closing_tag[2] === 2) {
2405 $trusted .= sq_sanitize(substr($body, $curpos, $closing_tag[4] - $curpos + 1),
2406 $orig_tag_list, $rm_tags_with_content, $self_closing_tags,
2407 $force_tag_closing, $rm_attnames, $bad_attvals, $add_attr_to_tag,
2408 $message, $id, $mailbox, true);
2409 $curpos = $closing_tag[4] + 1;
2410 continue 2;
2411 }
2412 $closing_tag_offset = $next_tag[0] + 1;
2413 }
2414 if ($closing_tag === false)
2415 { /* 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 */ }
2416 }
2417
691a2d25 2418 $free_content = substr($body, $curpos, $lt-$curpos);
2419 /**
8bd0068d 2420 * Take care of <style>
2421 */
7d06541f 2422 if ($tagname == "style" && $tagtype == 1){
da2415c1 2423 list($free_content, $curpos) =
e60a299a 2424 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
7d06541f 2425 if ($free_content != FALSE){
1d398b1d 2426 if ( !empty($attary) ) {
2427 $attary = sq_fixatts($tagname,
2428 $attary,
2429 $rm_attnames,
2430 $bad_attvals,
2431 $add_attr_to_tag,
2432 $message,
2433 $id,
2434 $mailbox
2435 );
2436 }
7d06541f 2437 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2438 $trusted .= $free_content;
2439 $trusted .= sq_tagprint($tagname, false, 2);
2440 }
2441 continue;
691a2d25 2442 }
2443 if ($skip_content == false){
2444 $trusted .= $free_content;
691a2d25 2445 }
2446 if ($tagname != FALSE){
2447 if ($tagtype == 2){
2448 if ($skip_content == $tagname){
2449 /**
8bd0068d 2450 * Got to the end of tag we needed to remove.
2451 */
691a2d25 2452 $tagname = false;
2453 $skip_content = false;
2454 } else {
2455 if ($skip_content == false){
c828931c 2456 if ($tagname == "body"){
2457 $tagname = "div";
2dd879b8 2458 }
da2415c1 2459 if (isset($open_tags{$tagname}) &&
8bd0068d 2460 $open_tags{$tagname} > 0){
2dd879b8 2461 $open_tags{$tagname}--;
691a2d25 2462 } else {
2dd879b8 2463 $tagname = false;
691a2d25 2464 }
691a2d25 2465 }
2466 }
2467 } else {
2468 /**
8bd0068d 2469 * $rm_tags_with_content
2470 */
691a2d25 2471 if ($skip_content == false){
2472 /**
8bd0068d 2473 * See if this is a self-closing type and change
2474 * tagtype appropriately.
2475 */
691a2d25 2476 if ($tagtype == 1
8bd0068d 2477 && in_array($tagname, $self_closing_tags)){
2dd879b8 2478 $tagtype = 3;
691a2d25 2479 }
2480 /**
8bd0068d 2481 * See if we should skip this tag and any content
2482 * inside it.
2483 */
691a2d25 2484 if ($tagtype == 1 &&
8bd0068d 2485 in_array($tagname, $rm_tags_with_content)){
691a2d25 2486 $skip_content = $tagname;
2487 } else {
da2415c1 2488 if (($rm_tags == false
8bd0068d 2489 && in_array($tagname, $tag_list)) ||
2490 ($rm_tags == true &&
2491 !in_array($tagname, $tag_list))){
691a2d25 2492 $tagname = false;
2493 } else {
2dd879b8 2494 /**
8bd0068d 2495 * Convert body into div.
2496 */
2dd879b8 2497 if ($tagname == "body"){
2498 $tagname = "div";
da2415c1 2499 $attary = sq_body2div($attary, $mailbox,
8bd0068d 2500 $message, $id);
2dd879b8 2501 }
691a2d25 2502 if ($tagtype == 1){
2503 if (isset($open_tags{$tagname})){
2504 $open_tags{$tagname}++;
2505 } else {
2506 $open_tags{$tagname}=1;
2507 }
2508 }
2509 /**
8bd0068d 2510 * This is where we run other checks.
2511 */
691a2d25 2512 if (is_array($attary) && sizeof($attary) > 0){
2513 $attary = sq_fixatts($tagname,
8bd0068d 2514 $attary,
2515 $rm_attnames,
2516 $bad_attvals,
2517 $add_attr_to_tag,
2518 $message,
2519 $id,
2520 $mailbox
2521 );
691a2d25 2522 }
2523 }
2524 }
691a2d25 2525 }
2526 }
2527 if ($tagname != false && $skip_content == false){
2528 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2529 }
691a2d25 2530 }
2531 $curpos = $gt+1;
a3daaaf3 2532 }
691a2d25 2533 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2534 if ($force_tag_closing == true){
2535 foreach ($open_tags as $tagname=>$opentimes){
2536 while ($opentimes > 0){
2537 $trusted .= '</' . $tagname . '>';
2538 $opentimes--;
2539 }
2540 }
2541 $trusted .= "\n";
2542 }
2543 $trusted .= "<!-- end sanitized html -->\n";
2544 return $trusted;
2545}
451f74a2 2546
691a2d25 2547/**
8bd0068d 2548 * This is a wrapper function to call html sanitizing routines.
2549 *
2550 * @param $body the body of the message
2551 * @param $id the id of the message
c189a963 2552
2553 * @param $message
2554 * @param $mailbox
2555 * @param boolean $take_mailto_links When TRUE, converts mailto: links
2556 * into internal SM compose links
2557 * (optional; default = TRUE)
8bd0068d 2558 * @return a string with html safe to display in the browser.
2559 */
c189a963 2560function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2561
202bcbcc 2562 // require_once(SM_PATH . 'functions/url_parser.php'); // for $MailTo_PReg_Match
c189a963 2563
691a2d25 2564 global $attachment_common_show_images, $view_unsafe_images,
dfc64d14 2565 $has_unsafe_images, $allow_svg_display, $rcdata_rawtext_tags,
2566 $remove_rcdata_rawtext_tags_and_content;
2567
2568 $rcdata_rawtext_tags = array(
2569 "noscript",
2570 "noframes",
2571 "noembed",
2572 "textarea",
2573 // also "title", "xmp", "script", "iframe", "plaintext" which we already remove below
2574 );
2575
691a2d25 2576 /**
8bd0068d 2577 * Don't display attached images in HTML mode.
2b665f28 2578 *
d0187bd6 2579 * SB: why?
8bd0068d 2580 */
691a2d25 2581 $attachment_common_show_images = false;
2582 $tag_list = Array(
dfc64d14 2583 false, // remove these tags
8bd0068d 2584 "meta",
2585 "html",
2586 "head",
2587 "base",
2588 "link",
2589 "frame",
2590 "iframe",
2591 "plaintext",
caf0ab1d 2592 "marquee",
8bd0068d 2593 );
691a2d25 2594
2595 $rm_tags_with_content = Array(
8bd0068d 2596 "script",
caf0ab1d 2597 "object",
8bd0068d 2598 "applet",
2599 "embed",
2600 "title",
2601 "frameset",
0493ed11 2602 "xmp",
caf0ab1d 2603 "xml",
8bd0068d 2604 );
c88cef21 2605 if (!$allow_svg_display)
caf0ab1d 2606 $rm_tags_with_content[] = 'svg';
dfc64d14 2607 /**
2608 * SquirrelMail will parse RCDATA and RAWTEXT tags and handle them as the special
2609 * case that they are, but if you prefer to remove them and their contents entirely
2610 * (in most cases, should be a safe thing with minimal impact), you can add the
2611 * following to config/config_local.php
2612 * $remove_rcdata_rawtext_tags_and_content = TRUE;
2613 */
2614 if ($remove_rcdata_rawtext_tags_and_content)
2615 $rm_tags_with_content = array_merge($rm_tags_with_content, $rcdata_rawtext_tags);
691a2d25 2616
2617 $self_closing_tags = Array(
8bd0068d 2618 "img",
2619 "br",
2620 "hr",
2621 "input",
caf0ab1d 2622 "outbind",
8bd0068d 2623 );
691a2d25 2624
2dd879b8 2625 $force_tag_closing = true;
691a2d25 2626
2627 $rm_attnames = Array(
8bd0068d 2628 "/.*/" =>
2629 Array(
2630 "/target/i",
2631 "/^on.*/i",
2632 "/^dynsrc/i",
2633 "/^data.*/i",
caf0ab1d 2634 "/^lowsrc.*/i",
8bd0068d 2635 )
2636 );
691a2d25 2637
6ab8321f 2638 global $use_transparent_security_image;
2639 if ($use_transparent_security_image) $secremoveimg = '../images/spacer.png';
2640 else $secremoveimg = '../images/' . _("sec_remove_eng.png");
2641
691a2d25 2642 $bad_attvals = Array(
8bd0068d 2643 "/.*/" =>
691a2d25 2644 Array(
0a6ec9b5 2645 "/^src|background/i" =>
8bd0068d 2646 Array(
691a2d25 2647 Array(
8bd0068d 2648 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2649 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2650 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
691a2d25 2651 ),
8bd0068d 2652 Array(
2653 "\\1$secremoveimg\\2",
2654 "\\1$secremoveimg\\2",
2655 "\\1$secremoveimg\\2",
8bd0068d 2656 )
2657 ),
0a6ec9b5 2658 "/^href|action/i" =>
8bd0068d 2659 Array(
0a6ec9b5 2660 Array(
8bd0068d 2661 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2662 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2663 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
0a6ec9b5 2664 ),
691a2d25 2665 Array(
8bd0068d 2666 "\\1#\\1",
2667 "\\1#\\1",
8bd0068d 2668 "\\1#\\1"
02474e43 2669 )
8bd0068d 2670 ),
2671 "/^style/i" =>
2672 Array(
2673 Array(
5db90261 2674 "/\/\*.*\*\//",
8bd0068d 2675 "/expression/i",
2676 "/binding/i",
2677 "/behaviou*r/i",
2678 "/include-source/i",
39352565 2679
2680 // position:relative can also be exploited
2681 // to put content outside of email body area
2682 // and position:fixed is similarly exploitable
2683 // as position:absolute, so we'll remove it
2684 // altogether....
2685 //
2686 // Does this screw up legitimate HTML messages?
2687 // If so, the only fix I see is to allow position
2688 // attributes (any values? I think we still have
2689 // to block static and fixed) only if $use_iframe
2690 // is enabled (1.5.0+)
2691 //
2692 // was: "/position\s*:\s*absolute/i",
2693 //
2694 "/position\s*:/i",
2695
1d935bc2 2696 "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
8bd0068d 2697 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2698 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2699 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
39352565 2700 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si",
8bd0068d 2701 ),
2702 Array(
5db90261 2703 "",
8bd0068d 2704 "idiocy",
2705 "idiocy",
2706 "idiocy",
2707 "idiocy",
567dc524 2708 "idiocy",
5b4884be 2709 "url",
8bd0068d 2710 "url(\\1#\\1)",
2711 "url(\\1#\\1)",
2712 "url(\\1#\\1)",
8bd0068d 2713 "\\1:url(\\2#\\3)"
2714 )
691a2d25 2715 )
8bd0068d 2716 )
691a2d25 2717 );
9ebace19 2718
d31f73f1 2719 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
2720 // images off by default.
afbf184c 2721 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET, FALSE);
9ebace19 2722
691a2d25 2723 if (!$view_unsafe_images){
2724 /**
8bd0068d 2725 * Remove any references to http/https if view_unsafe_images set
2726 * to false.
2727 */
02474e43 2728 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
8bd0068d 2729 '/^([\'\"])\s*https*:.*([\'\"])/si');
02474e43 2730 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
8bd0068d 2731 "\\1$secremoveimg\\1");
02474e43 2732 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
7ef0b415 2733 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
02474e43 2734 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
8bd0068d 2735 "url(\\1$secremoveimg\\1)");
691a2d25 2736 }
451f74a2 2737
691a2d25 2738 $add_attr_to_tag = Array(
8bd0068d 2739 "/^a$/i" =>
c25f2fbb 2740 Array('target'=>'"_blank"',
02474e43 2741 'title'=>'"'._("This external link will open in a new window").'"'
8bd0068d 2742 )
2743 );
da2415c1 2744 $trusted = sq_sanitize($body,
8bd0068d 2745 $tag_list,
2746 $rm_tags_with_content,
2747 $self_closing_tags,
2748 $force_tag_closing,
2749 $rm_attnames,
2750 $bad_attvals,
2751 $add_attr_to_tag,
2752 $message,
2753 $id,
2754 $mailbox
2755 );
567dc524 2756 if (strpos($trusted,$secremoveimg)){
691a2d25 2757 $has_unsafe_images = true;
da2415c1 2758 }
c189a963 2759
2760 // we want to parse mailto's in HTML output, change to SM compose links
2761 // this is a modified version of code from url_parser.php... but Marc is
2762 // right: we need a better filtering implementation; adding this randomly
2763 // here is not a great solution
2764 //
2765 if ($take_mailto_links) {
2766 // parseUrl($trusted); // this even parses URLs inside of tags... too aggressive
2767 global $MailTo_PReg_Match;
7196da62 2768 // some mailers (Microsoft, surprise surprise) produce mailto strings without being
2769 // inside an anchor (link) tag, so we have to make sure the regex looks for the
2770 // quote before mailto, and we'll also try to convert the non-links back into links
2771 $MailTo_PReg_Match = '/([\'"])?mailto:' . substr($MailTo_PReg_Match, 1) ;
c189a963 2772 if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2773 foreach ($regs[0] as $i => $mailto_before) {
7196da62 2774 $mailto_params = $regs[11][$i];
2775
2776 // get rid of any leading quote we may have captured but don't care about
2777 //
2778 $mailto_before = ltrim($mailto_before, '"\'');
2779
c189a963 2780 // get rid of any tailing quote since we have to add send_to to the end
2781 //
7196da62 2782 $mailto_before = rtrim($mailto_before, '"\'');
2783 $mailto_params = rtrim($mailto_params, '"\'');
c189a963 2784
7196da62 2785 if ($regs[2][$i]) { //if there is an email addr before '?', we need to merge it with the params
2786 $to = 'to=' . $regs[2][$i];
c189a963 2787 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
2788 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2789 else {
2790 if ($mailto_params) //already some params, append to them
2791 $mailto_params .= '&amp;' . $to;
2792 else
2793 $mailto_params .= '?' . $to;
2794 }
2795 }
2796
2797 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2798
2799 // we'll already have target=_blank, no need to allow comp_in_new
2800 // here (which would be a lot more work anyway)
2801 //
2802 global $compose_new_win;
2803 $temp_comp_in_new = $compose_new_win;
2804 $compose_new_win = 0;
2805 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2806 $compose_new_win = $temp_comp_in_new;
2807
2808 // remove <a href=" and anything after the next quote (we only
2809 // need the uri, not the link HTML) in compose uri
2810 //
7196da62 2811 // but only do this if the original mailto was in a real anchor tag
2812 //
2813 if (!empty($regs[1][$i])) {
2814 $comp_uri = substr($comp_uri, 9);
2815 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2816 }
c189a963 2817 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2818 }
2819 }
2820 }
2821
691a2d25 2822 return $trusted;
451f74a2 2823}
a4a70693 2824
da2415c1 2825/**
8bd0068d 2826 * function SendDownloadHeaders - send file to the browser
2827 *
2828 * Original Source: SM core src/download.php
2829 * moved here to make it available to other code, and separate
2830 * front end from back end functionality.
2831 *
2832 * @param string $type0 first half of mime type
2833 * @param string $type1 second half of mime type
2834 * @param string $filename filename to tell the browser for downloaded file
2835 * @param boolean $force whether to force the download dialog to pop
2836 * @param optional integer $filesize send the Content-Header and length to the browser
2837 * @return void
2838 */
02474e43 2839function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2840 global $languages, $squirrelmail_language;
cfffd60b 2841 $isIE = $isIE6plus = false;
02474e43 2842
2843 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2844
2845 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
8bd0068d 2846 strstr($HTTP_USER_AGENT, 'Opera') === false) {
cfffd60b 2847 $isIE = true;
02474e43 2848 }
2849
cfffd60b 2850 if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2851 ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2852 $isIE6plus = true;
02474e43 2853 }
2854
2855 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 2856 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
02474e43 2857 $filename =
8bd0068d 2858 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
02474e43 2859 } else {
1a4bc4a6 2860 $filename = preg_replace('/[\\\\\/:*?"<>|;]/', '_', str_replace('&nbsp;', ' ', $filename));
02474e43 2861 }
2862
2863 // A Pox on Microsoft and it's Internet Explorer!
2864 //
2865 // IE has lots of bugs with file downloads.
2866 // It also has problems with SSL. Both of these cause problems
2867 // for us in this function.
2868 //
2869 // See this article on Cache Control headers and SSL
2870 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2871 //
2872 // The best thing you can do for IE is to upgrade to the latest
2873 // version
2874 //set all the Cache Control Headers for IE
2875 if ($isIE) {
2876 $filename=rawurlencode($filename);
2877 header ("Pragma: public");
8bd0068d 2878 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
c2503a4e 2879 // does nothing - see: https://blogs.msdn.microsoft.com/ieinternals/2009/07/20/internet-explorers-cache-control-extensions/
2880 // header ("Cache-Control: post-check=0, pre-check=0", false);
8bd0068d 2881 header ("Cache-Control: private");
02474e43 2882
2883 //set the inline header for IE, we'll add the attachment header later if we need it
2884 header ("Content-Disposition: inline; filename=$filename");
2885 }
2886
2887 if (!$force) {
2888 // Try to show in browser window
2889 header ("Content-Disposition: inline; filename=\"$filename\"");
2890 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2891 } else {
2892 // Try to pop up the "save as" box
2893
2894 // IE makes this hard. It pops up 2 save boxes, or none.
2895 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2896 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2897 // But, according to Microsoft, it is "RFC compliant but doesn't
2898 // take into account some deviations that allowed within the
2899 // specification." Doesn't that mean RFC non-compliant?
2900 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2901
2902 // all browsers need the application/octet-stream header for this
2903 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2904
2905 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2906 // Do not have quotes around filename, but that applied to
2907 // "attachment"... does it apply to inline too?
2908 header ("Content-Disposition: attachment; filename=\"$filename\"");
2909
cfffd60b 2910 if ($isIE && !$isIE6plus) {
02474e43 2911 // This combination seems to work mostly. IE 5.5 SP 1 has
2912 // known issues (see the Microsoft Knowledge Base)
2913
2914 // This works for most types, but doesn't work with Word files
2915 header ("Content-Type: application/download; name=\"$filename\"");
7e2ff844 2916 header ("Content-Type: application/force-download; name=\"$filename\"");
02474e43 2917 // These are spares, just in case. :-)
2918 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2919 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2920 //header("Content-Type: application/octet-stream; name=\"$filename\"");
7e2ff844 2921 } else if ($isIE) {
2922 // This is to prevent IE for MIME sniffing and auto open a file in IE
2923 header ("Content-Type: application/force-download; name=\"$filename\"");
02474e43 2924 } else {
2925 // another application/octet-stream forces download for Netscape
2926 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2927 }
2928 }
2929
2930 //send the content-length header if the calling function provides it
2931 if ($filesize > 0) {
2932 header("Content-Length: $filesize");
2933 }
07c49f57 2934
8d863f64 2935} // end fn SendDownloadHeaders