Fix improperly quoted href link addresses; closes XSS exploit exlained at CVE-2008...
[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 *
4b5049de 9 * @copyright &copy; 1999-2007 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 />';
472e7acb 66 $errormessage .= '<pre>' . htmlspecialchars($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 $msg->is_flagged = true;
94 }
95 break;
96 case 'M':
97 if (strtolower($flag) == '$mdnsent') {
98 $msg->is_mdnsent = true;
99 }
100 break;
101 default:
102 break;
7a9e9c89 103 }
104 }
451f74a2 105 }
7a9e9c89 106 // listEntities($msg);
3d8371be 107 return $msg;
451f74a2 108}
b74ba498 109
22efa9fb 110
111
3d8371be 112/* This starts the parsing of a particular structure. It is called recursively,
8bd0068d 113 * so it can be passed different structures. It returns an object of type
114 * $message.
115 * First, it checks to see if it is a multipart message. If it is, then it
116 * handles that as it sees is necessary. If it is just a regular entity,
117 * then it parses it and adds the necessary header information (by calling out
118 * to mime_get_elements()
119 */
451f74a2 120
4d592352 121function mime_fetch_body($imap_stream, $id, $ent_id=1, $fetch_size=0) {
3d8371be 122 /* Do a bit of error correction. If we couldn't find the entity id, just guess
8bd0068d 123 * that it is the first one. That is usually the case anyway.
124 */
7c7b74b3 125
09a4bde3 126 if (!$ent_id) {
08b7f7cc 127 $cmd = "FETCH $id BODY[]";
1035e159 128 } else {
08b7f7cc 129 $cmd = "FETCH $id BODY[$ent_id]";
09a4bde3 130 }
3d8371be 131
4d592352 132 if ($fetch_size!=0) $cmd .= "<0.$fetch_size>";
da2415c1 133
6201339c 134 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, TRUE);
77b88425 135 do {
3d8371be 136 $topline = trim(array_shift($data));
137 } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
a4a70693 138
451f74a2 139 $wholemessage = implode('', $data);
140 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
3d8371be 141 $ret = substr($wholemessage, 0, $regs[1]);
142 /* There is some information in the content info header that could be important
8bd0068d 143 * in order to parse html messages. Let's get them here.
144 */
0600bdf1 145// if ($ret{0} == '<') {
6201339c 146// $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, TRUE);
0600bdf1 147// }
451f74a2 148 } else if (ereg('"([^"]*)"', $topline, $regs)) {
149 $ret = $regs[1];
06bcb9c3 150 } else if ((stristr($topline, 'nil') !== false) && (empty($wholemessage))) {
151 $ret = $wholemessage;
451f74a2 152 } else {
153 global $where, $what, $mailbox, $passed_id, $startMessage;
3d8371be 154 $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
451f74a2 155 if (isset($where) && isset($what)) {
3d8371be 156 $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
a3daaaf3 157 } else {
3d8371be 158 $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
451f74a2 159 }
e5ea9327 160 $par .= '&amp;response=' . urlencode($response) .
8bd0068d 161 '&amp;message=' . urlencode($message) .
162 '&amp;topline=' . urlencode($topline);
a019eeb8 163
8bd0068d 164 echo '<tt><br />' .
02474e43 165 '<table width="80%"><tr>' .
166 '<tr><td colspan="2">' .
167 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
168 '</td></tr>' .
169 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
170 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
171 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
172 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
173 "</table><br /></tt></font><hr />";
346817d4 174
6201339c 175 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, TRUE);
451f74a2 176 array_shift($data);
177 $wholemessage = implode('', $data);
a019eeb8 178
346817d4 179 $ret = $wholemessage;
a3daaaf3 180 }
3d8371be 181 return $ret;
451f74a2 182}
d4467150 183
6cc670de 184function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding, $rStream='php://stdout') {
1035e159 185
3d8371be 186 /* Don't kill the connection if the browser is over a dialup
8bd0068d 187 * and it would take over 30 seconds to download it.
188 * Don't call set_time_limit in safe mode.
189 */
b7206e1d 190
3d8371be 191 if (!ini_get('safe_mode')) {
b7206e1d 192 set_time_limit(0);
193 }
7c7b74b3 194 /* in case of base64 encoded attachments, do not buffer them.
8bd0068d 195 Instead, echo the decoded attachment directly to screen */
7c7b74b3 196 if (strtolower($encoding) == 'base64') {
197 if (!$ent_id) {
7591f143 198 $query = "FETCH $id BODY[]";
7c7b74b3 199 } else {
7591f143 200 $query = "FETCH $id BODY[$ent_id]";
7c7b74b3 201 }
7591f143 202 sqimap_run_command($imap_stream,$query,true,$response,$message,TRUE,'sqimap_base64_decode',$rStream,true);
1d142b8d 203 } else {
7591f143 204 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
205 if (is_resource($rStream)) {
206 fputs($rStream,decodeBody($body,$encoding));
207 } else {
208 echo decodeBody($body, $encoding);
209 }
1d142b8d 210 }
346817d4 211
da2415c1 212 /*
8bd0068d 213 TODO, use the same method for quoted printable.
214 However, I assume that quoted printable attachments aren't that large
215 so the performancegain / memory usage drop will be minimal.
216 If we decide to add that then we need to adapt sqimap_fread because
217 we need to split te result on \n and fread doesn't stop at \n. That
218 means we also should provide $results from sqimap_fread (by ref) to
219 te function and set $no_return to false. The $filter function for
220 quoted printable should handle unsetting of $results.
221 */
da2415c1 222 /*
8bd0068d 223 TODO 2: find out how we write to the output stream php://stdout. fwrite
224 doesn't work because 'php://stdout isn't a stream.
225 */
7c7b74b3 226
5d9c6f73 227 return;
451f74a2 228}
beb9e459 229
451f74a2 230/* -[ END MIME DECODING ]----------------------------------------------------------- */
d4467150 231
3d8371be 232/* This is here for debugging purposes. It will print out a list
8bd0068d 233 * of all the entity IDs that are in the $message object.
234 */
451f74a2 235function listEntities ($message) {
3d8371be 236 if ($message) {
3c621ba1 237 echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br />';
3d8371be 238 for ($i = 0; isset($message->entities[$i]); $i++) {
239 echo "$i : ";
240 $msg = listEntities($message->entities[$i]);
241
242 if ($msg) {
243 echo "return: ";
244 return $msg;
245 }
246 }
a4a70693 247 }
451f74a2 248}
f0c4dc12 249
f792c641 250function getPriorityStr($priority) {
3d8371be 251 $priority_level = substr($priority,0,1);
252
253 switch($priority_level) {
254 /* Check for a higher then normal priority. */
255 case '1':
256 case '2':
257 $priority_string = _("High");
258 break;
259
260 /* Check for a lower then normal priority. */
261 case '4':
262 case '5':
263 $priority_string = _("Low");
264 break;
265
266 /* Check for a normal priority. */
267 case '3':
268 default:
269 $priority_level = '3';
270 $priority_string = _("Normal");
271 break;
272
273 }
274 return $priority_string;
f792c641 275}
276
451f74a2 277/* returns a $message object for a particular entity id */
278function getEntity ($message, $ent_id) {
a4a70693 279 return $message->getEntity($ent_id);
451f74a2 280}
8beafbbc 281
3d8371be 282/* translateText
8bd0068d 283 * Extracted from strings.php 23/03/2002
284 */
da4c66e8 285
286function translateText(&$body, $wrap_at, $charset) {
3d8371be 287 global $where, $what; /* from searching */
288 global $color; /* color theme */
da4c66e8 289
202bcbcc 290 // require_once(SM_PATH . 'functions/url_parser.php');
da4c66e8 291
292 $body_ary = explode("\n", $body);
da4c66e8 293 for ($i=0; $i < count($body_ary); $i++) {
294 $line = $body_ary[$i];
295 if (strlen($line) - 2 >= $wrap_at) {
c7aff938 296 sqWordWrap($line, $wrap_at, $charset);
da4c66e8 297 }
298 $line = charset_decode($charset, $line);
299 $line = str_replace("\t", ' ', $line);
300
301 parseUrl ($line);
302
3d8371be 303 $quotes = 0;
da4c66e8 304 $pos = 0;
3d8371be 305 $j = strlen($line);
da4c66e8 306
3d8371be 307 while ($pos < $j) {
da4c66e8 308 if ($line[$pos] == ' ') {
3d8371be 309 $pos++;
da4c66e8 310 } else if (strpos($line, '&gt;', $pos) === $pos) {
311 $pos += 4;
3d8371be 312 $quotes++;
da4c66e8 313 } else {
314 break;
315 }
316 }
3d8371be 317
83c94382 318 if ($quotes % 2) {
d0814c02 319 $line = '<span class="quote1">' . $line . '</span>';
4c25967c 320 } elseif ($quotes) {
d0814c02 321 $line = '<span class="quote2">' . $line . '</span>';
da4c66e8 322 }
3d8371be 323
da4c66e8 324 $body_ary[$i] = $line;
325 }
326 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
327}
328
a2bfcbce 329/**
da1b55ad 330 * This returns a parsed string called $body. That string can then
331 * be displayed as the actual message in the HTML. It contains
332 * everything needed, including HTML Tags, Attachments at the
333 * bottom, etc.
f8a1ed5a 334 *
da1b55ad 335 * Since 1.2.0 function uses message_body hook.
336 * Till 1.3.0 function included output of formatAttachments().
337 *
338 * @param resource $imap_stream imap connection resource
339 * @param object $message squirrelmail message object
340 * @param array $color squirrelmail color theme array
341 * @param integer $wrap_at number of characters per line
342 * @param string $ent_num (since 1.3.0) message part id
343 * @param integer $id (since 1.3.0) message id
344 * @param string $mailbox (since 1.3.0) imap folder name
da1b55ad 345 * @return string html formated message text
346 */
a540f994 347function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
3d8371be 348 /* This if statement checks for the entity to show as the
8bd0068d 349 * primary message. To add more of them, just put them in the
350 * order that is their priority.
351 */
ce68b76b 352 global $startMessage, $languages, $squirrelmail_language,
40a34e57 353 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id,
14c85e39 354 $use_iframe, $iframe_height, $download_and_unsafe_link,
955bfc8f 355 $download_href, $unsafe_image_toggle_href, $unsafe_image_toggle_text,
551c7b53 356 $oTemplate, $nbsp;
2c25d36a 357
358 // workaround for not updated config.php
359 if (! isset($use_iframe)) $use_iframe = false;
77bfbd2e 360
d31f73f1 361 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
362 // images off by default.
afbf184c 363 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET, FALSE);
d03c24f4 364
cc34b00d 365 $body = '';
23bcec6f 366 $urlmailbox = urlencode($mailbox);
451f74a2 367 $body_message = getEntity($message, $ent_num);
368 if (($body_message->header->type0 == 'text') ||
8bd0068d 369 ($body_message->header->type0 == 'rfc822')) {
3d8371be 370 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
451f74a2 371 $body = decodeBody($body, $body_message->header->encoding);
e842b215 372
373 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 374 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
e842b215 375 if (mb_detect_encoding($body) != 'ASCII') {
33a55f5a 376 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
e842b215 377 }
378 }
d849b570 379
380 /* As of 1.5.2, $body is passed (and modified) by reference */
381 do_hook('message_body', $body);
23bcec6f 382
3d8371be 383 /* If there are other types that shouldn't be formatted, add
8bd0068d 384 * them here.
385 */
3d8371be 386
451f74a2 387 if ($body_message->header->type1 == 'html') {
3d8371be 388 if ($show_html_default <> 1) {
85015544 389 $entity_conv = array('&nbsp;' => ' ',
8bd0068d 390 '<p>' => "\n",
391 '<P>' => "\n",
392 '<br>' => "\n",
393 '<BR>' => "\n",
394 '<br />' => "\n",
395 '<BR />' => "\n",
396 '&gt;' => '>',
397 '&lt;' => '<');
85015544 398 $body = strtr($body, $entity_conv);
3d8371be 399 $body = strip_tags($body);
85015544 400 $body = trim($body);
401 translateText($body, $wrap_at,
8bd0068d 402 $body_message->header->getParameter('charset'));
a540f994 403 } elseif ($use_iframe) {
84410f31 404 /**
405 * If we don't add html message between iframe tags,
406 * we must detect unsafe images and modify $has_unsafe_images.
f8a1ed5a 407 */
758a7889 408 $html_body = magicHTML($body, $id, $message, $mailbox);
b6c52e61 409 // Convert character set in order to display html mails in different character set
410 $html_body = charset_decode($body_message->header->getParameter('charset'),$html_body,false,true);
84410f31 411
2c25d36a 412 // creating iframe url
413 $iframeurl=sqm_baseuri().'src/view_html.php?'
f8a1ed5a 414 . 'mailbox=' . $urlmailbox
2c25d36a 415 . '&amp;passed_id=' . $id
416 . '&amp;ent_id=' . $ent_num
417 . '&amp;view_unsafe_images=' . (int) $view_unsafe_images;
418
79d58d4c 419 global $oTemplate;
420 $oTemplate->assign('iframe_url', $iframeurl);
51554708 421 $oTemplate->assign('iframe_height', $iframe_height);
79d58d4c 422 $oTemplate->assign('html_body', $html_body);
2b665f28 423
79d58d4c 424 $body = $oTemplate->fetch('read_html_iframe.tpl');
a3daaaf3 425 } else {
2c25d36a 426 // old way of html rendering
b6c52e61 427 /**
758a7889 428 * convert character set. charset_decode does not remove html special chars
b6c52e61 429 * applied by magicHTML functions and does not sanitize them second time if
758a7889 430 * fourth argument is true.
431 */
567dc524 432 $charset = $body_message->header->getParameter('charset');
433 if (!empty($charset)) {
434 $body = charset_decode($charset,$body,false,true);
435 }
436 $body = magicHTML($body, $id, $message, $mailbox);
a3daaaf3 437 }
451f74a2 438 } else {
3d8371be 439 translateText($body, $wrap_at,
8bd0068d 440 $body_message->header->getParameter('charset'));
451f74a2 441 }
a2bfcbce 442
9ebace19 443 /*
444 * Previously the links for downloading and unsafe images were printed
445 * under the mail. By putting the links in a global variable we can
446 * print it in the toolbar where it belongs. Since the original code was
447 * in this place it's left here. It might be possible to move it to some
448 * other place if that makes sense. The possibility to do so has not
449 * been evaluated yet.
450 */
451
452 // Initialize the global variable to an empty string.
453 // FIXME: To have $download_and_unsafe_link as a global variable might not be needed since the use of separate variables ($download_href, $unsafe_image_toggle_href, and $unsafe_image_toggle_text) for the templates was introduced.
40a34e57 454 $download_and_unsafe_link = '';
455
9ebace19 456 // Prepare and build a link for downloading the mail.
83cf04bd 457 $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
8bd0068d 458 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
459 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
08b7f7cc 460 if (isset($passed_ent_id)) {
461 $link .= '&amp;passed_ent_id='.$passed_ent_id;
462 }
14c85e39 463 $download_href = SM_PATH . 'src/download.php?absolute_dl=true&amp;' . $link;
9ebace19 464
465 // Always add the link for downloading the mail as a file to the global
466 // variable.
955bfc8f 467 $download_and_unsafe_link .= "$nbsp|$nbsp"
468 . create_hyperlink($download_href, _("Download this as a file"));
9ebace19 469
470 // Find out the right text to use in the link depending on the
471 // circumstances. If the unsafe images are displayed the link should
472 // hide them, if they aren't displayed the link should only appear if
473 // the mail really contains unsafe images.
7aad7b77 474 if ($view_unsafe_images) {
23f617b8 475 $text = _("Hide Unsafe Images");
7aad7b77 476 } else {
08b7f7cc 477 if (isset($has_unsafe_images) && $has_unsafe_images) {
478 $link .= '&amp;view_unsafe_images=1';
479 $text = _("View Unsafe Images");
480 } else {
481 $text = '';
482 }
3d8371be 483 }
9ebace19 484
485 // Only create a link for unsafe images if there's need for one. If so:
486 // add it to the global variable.
83cf04bd 487 if($text != '') {
14c85e39 488 $unsafe_image_toggle_href = SM_PATH . 'src/read_body.php?'.$link;
489 $unsafe_image_toggle_text = $text;
955bfc8f 490 $download_and_unsafe_link .= "$nbsp|$nbsp"
491 . create_hyperlink($unsafe_image_toggle_href, $text);
83cf04bd 492 }
3d8371be 493 }
494 return $body;
451f74a2 495}
b74ba498 496
da1b55ad 497/**
a540f994 498 * Generate attachments array for passing to templates.
2b665f28 499 *
d67f519a 500 * @since 1.5.2
da1b55ad 501 * @param object $message SquirrelMail message object
502 * @param array $exclude_id message parts that are not attachments.
503 * @param string $mailbox mailbox name
504 * @param integer $id message id
da1b55ad 505 */
d67f519a 506function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
507 global $where, $what, $startMessage, $color, $passed_ent_id, $base_uri;
451f74a2 508
23bcec6f 509 $att_ar = $message->getAttachments($exclude_id);
23bcec6f 510 $urlMailbox = urlencode($mailbox);
511
d67f519a 512 $attachments = array();
23bcec6f 513 foreach ($att_ar as $att) {
fdc9d9b5 514 $ent = $att->entity_id;
2e25760a 515 $header = $att->header;
f0c4dc12 516 $type0 = strtolower($header->type0);
517 $type1 = strtolower($header->type1);
2e25760a 518 $name = '';
d0187bd6 519 $links = array();
21dab2dc 520 $links['download link']['text'] = _("Download");
202bcbcc 521 $links['download link']['href'] = $base_uri .
8bd0068d 522 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
2b665f28 523
2e25760a 524 if ($type0 =='message' && $type1 == 'rfc822') {
202bcbcc 525 $default_page = $base_uri . 'src/read_body.php';
2e25760a 526 $rfc822_header = $att->rfc822_header;
098ea084 527 $filename = $rfc822_header->subject;
6cc08d8b 528 if (trim( $filename ) == '') {
529 $filename = 'untitled-[' . $ent . ']' ;
08b7f7cc 530 }
2e25760a 531 $from_o = $rfc822_header->from;
532 if (is_object($from_o)) {
04ea844e 533 $from_name = decodeHeader($from_o->getAddress(false));
7e697748 534 } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
535 // something weird happens when a digest message is opened and you return to the digest
536 // now the from object is part of an array. Probably the parseHeader call overwrites the info
537 // retrieved from the bodystructure in a different way. We need to fix this later.
538 // possible starting point, do not fetch header we already have and inspect how
539 // the rfc822_header object behaves.
540 $from_name = decodeHeader($from_o[0]->getAddress(false));
2e25760a 541 } else {
542 $from_name = _("Unknown sender");
f0c4dc12 543 }
d0187bd6 544 $description = _("From").': '.$from_name;
23bcec6f 545 } else {
202bcbcc 546 $default_page = $base_uri . 'src/download.php';
02474e43 547 $filename = $att->getFilename();
f810c0b2 548 if ($header->description) {
098ea084 549 $description = decodeHeader($header->description);
f810c0b2 550 } else {
3d8371be 551 $description = '';
552 }
2e25760a 553 }
554
555 $display_filename = $filename;
3d8371be 556 if (isset($passed_ent_id)) {
557 $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
558 } else {
559 $passed_ent_id_link = '';
560 }
561 $defaultlink = $default_page . "?startMessage=$startMessage"
8bd0068d 562 . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
563 . '&amp;ent_id='.$ent.$passed_ent_id_link;
2e25760a 564 if ($where && $what) {
8bd0068d 565 $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
2e25760a 566 }
7e2ff844 567 // IE does make use of mime content sniffing. Forcing a download
568 // prohibit execution of XSS inside an application/octet-stream attachment
569 if ($type0 == 'application' && $type1 == 'octet-stream') {
570 $defaultlink .= '&amp;absolute_dl=true';
571 }
2b665f28 572
3d8371be 573 /* This executes the attachment hook with a specific MIME-type.
8bd0068d 574 * If that doesn't have results, it tries if there's a rule
9b94c54d 575 * for a more generic type. Finally, a hook for ALL attachment
576 * types is run as well.
8bd0068d 577 */
8dca4d22 578 // First remember the default link.
579 $defaultlink_orig = $defaultlink;
580
d849b570 581 /* The API for this hook has changed as of 1.5.2 so that all plugin
582 arguments are passed in an array instead of each their own plugin
583 argument, and arguments are passed by reference, so instead of
584 returning any changes, changes should simply be made to the original
585 arguments themselves. */
9c3b2d22 586 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
587 &$defaultlink, &$display_filename, &$where, &$what);
588 do_hook("attachment $type0/$type1", $temp);
8dca4d22 589 if(count($links) <= 1 && $defaultlink == $defaultlink_orig) {
d849b570 590 /* The API for this hook has changed as of 1.5.2 so that all plugin
591 arguments are passed in an array instead of each their own plugin
592 argument, and arguments are passed by reference, so instead of
593 returning any changes, changes should simply be made to the original
594 arguments themselves. */
9c3b2d22 595 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
596 &$defaultlink, &$display_filename, &$where, &$what);
597 do_hook("attachment $type0/*", $temp);
2e25760a 598 }
d849b570 599 /* The API for this hook has changed as of 1.5.2 so that all plugin
600 arguments are passed in an array instead of each their own plugin
601 argument, and arguments are passed by reference, so instead of
602 returning any changes, changes should simply be made to the original
603 arguments themselves. */
9c3b2d22 604 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
605 &$defaultlink, &$display_filename, &$where, &$what);
8dca4d22 606 // Do not let a generic plugin change the default link if a more
607 // specialized one already did it...
608 if ($defaultlink != $defaultlink_orig) {
609 $dummy = '';
610 $temp[5] = &$dummy;
611 }
9c3b2d22 612 do_hook("attachment */*", $temp);
77b88425 613
d67f519a 614 $this_attachment = array();
615 $this_attachment['Name'] = decodeHeader($display_filename);
616 $this_attachment['Description'] = $description;
617 $this_attachment['DefaultHREF'] = $defaultlink;
618 $this_attachment['DownloadHREF'] = $links['download link']['href'];
619 $this_attachment['ViewHREF'] = isset($links['attachment_common']) ? $links['attachment_common']['href'] : '';
620 $this_attachment['Size'] = $header->size;
621 $this_attachment['ContentType'] = htmlspecialchars($type0 .'/'. $type1);
622 $this_attachment['OtherLinks'] = array();
3d8371be 623 foreach ($links as $val) {
d0187bd6 624 if ($val['text']==_("Download") || $val['text'] == _("View"))
625 continue;
626 if (empty($val['text']) && empty($val['extra']))
627 continue;
2b665f28 628
d67f519a 629 $temp = array();
630 $temp['HREF'] = $val['href'];
631 $temp['Text'] = (empty($val['text']) ? '' : $val['text']) . (empty($val['extra']) ? '' : $val['extra']);
632 $this_attachment['OtherLinks'][] = $temp;
2e25760a 633 }
d67f519a 634 $attachments[] = $this_attachment;
2b665f28 635
3d8371be 636 unset($links);
2e25760a 637 }
2b665f28 638
d67f519a 639 return $attachments;
640}
641
642/**
643 * Displays attachment links and information
644 *
645 * Since 1.3.0 function is not included in formatBody() call.
646 *
647 * Since 1.0.2 uses attachment $type0/$type1 hook.
648 * Since 1.2.5 uses attachment $type0/* hook.
649 * Since 1.5.0 uses attachments_bottom hook.
650 * Since 1.5.2 uses templates and does *not* return a value.
651 *
652 * @param object $message SquirrelMail message object
653 * @param array $exclude_id message parts that are not attachments.
654 * @param string $mailbox mailbox name
655 * @param integer $id message id
656 */
657function formatAttachments($message, $exclude_id, $mailbox, $id) {
658 global $oTemplate;
2b665f28 659
d67f519a 660 $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
d0187bd6 661
662 $oTemplate->assign('attachments', $attach);
663 $oTemplate->display('read_attachments.tpl');
451f74a2 664}
b74ba498 665
7c7b74b3 666function sqimap_base64_decode(&$string) {
7c0ec1d8 667
b17a8968 668 // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
7c0ec1d8 669 // fly decoding (to reduce memory usage) you have to check if the
670 // data has incomplete pairs
671
b17a8968 672 // Remove the noise in order to check if the 4 bytes pairs are complete
7c0ec1d8 673 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
674
42ce44f8 675 $sStringRem = '';
7c0ec1d8 676 $iMod = strlen($string) % 4;
677 if ($iMod) {
678 $sStringRem = substr($string,-$iMod);
b17a8968 679 // Check if $sStringRem contains padding characters
7c0ec1d8 680 if (substr($sStringRem,-1) != '=') {
681 $string = substr($string,0,-$iMod);
682 } else {
683 $sStringRem = '';
684 }
685 }
7c7b74b3 686 $string = base64_decode($string);
7c0ec1d8 687 return $sStringRem;
7c7b74b3 688}
689
fdf7cef1 690/**
691 * Decodes encoded message body
692 *
693 * This function decodes the body depending on the encoding type.
694 * Currently quoted-printable and base64 encodings are supported.
695 * decode_body hook was added to this function in 1.4.2/1.5.0
696 * @param string $body encoded message body
697 * @param string $encoding used encoding
698 * @return string decoded string
699 * @since 1.0
700 */
451f74a2 701function decodeBody($body, $encoding) {
83be314a 702
b583c3e8 703 $body = str_replace("\r\n", "\n", $body);
704 $encoding = strtolower($encoding);
3d8371be 705
d849b570 706 $encoding_handler = do_hook('decode_body', $encoding);
5166f86a 707
708
709 // plugins get first shot at decoding the body
710 //
711 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
712 $body = $encoding_handler('decode', $body);
713
fdf7cef1 714 } elseif ($encoding == 'quoted-printable' ||
8bd0068d 715 $encoding == 'quoted_printable') {
fdf7cef1 716 /**
717 * quoted_printable_decode() function is broken in older
718 * php versions. Text with \r\n decoding was fixed only
758a7889 719 * in php 4.3.0. Minimal code requirement 4.0.4 +
fdf7cef1 720 * str_replace("\r\n", "\n", $body); call.
721 */
b583c3e8 722 $body = quoted_printable_decode($body);
fdf7cef1 723 } elseif ($encoding == 'base64') {
b583c3e8 724 $body = base64_decode($body);
725 }
3d8371be 726
b583c3e8 727 // All other encodings are returned raw.
3d8371be 728 return $body;
451f74a2 729}
730
9f7f68c3 731/**
8bd0068d 732 * Decodes headers
733 *
e1115979 734 * This function decodes strings that are encoded according to
8bd0068d 735 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
736 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
737 *
738 * @param string $string header string that has to be made readable
739 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
bcae324b 740 * @param boolean $htmlsafe preserve spaces and sanitize html special characters. defaults to true
8bd0068d 741 * @param boolean $decide decide if string can be utfencoded. defaults to false
742 * @return string decoded header string
743 */
bcae324b 744function decodeHeader ($string, $utfencode=true,$htmlsafe=true,$decide=false) {
d6f584fc 745 global $languages, $squirrelmail_language,$default_charset;
79e07c7e 746 if (is_array($string)) {
747 $string = implode("\n", $string);
748 }
da2415c1 749
10dec454 750 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 751 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
33a55f5a 752 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
08b7f7cc 753 // Do we need to return at this point?
754 // return $string;
83be314a 755 }
79e07c7e 756 $i = 0;
08b7f7cc 757 $iLastMatch = -2;
db65b6b0 758 $encoded = true;
0a06275a 759
098ea084 760 $aString = explode(' ',$string);
08b7f7cc 761 $ret = '';
098ea084 762 foreach ($aString as $chunk) {
358a78a1 763 if ($encoded && $chunk === '') {
08b7f7cc 764 continue;
358a78a1 765 } elseif ($chunk === '') {
08b7f7cc 766 $ret .= ' ';
767 continue;
768 }
098ea084 769 $encoded = false;
08b7f7cc 770 /* if encoded words are not separated by a linear-space-white we still catch them */
771 $j = $i-1;
7e6ca3e8 772
08b7f7cc 773 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
774 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
775 if ($iLastMatch !== $j) {
bcae324b 776 if ($htmlsafe) {
9f7f68c3 777 $ret .= '&#32;';
08b7f7cc 778 } else {
779 $ret .= ' ';
780 }
781 }
782 $iLastMatch = $i;
783 $j = $i;
bcae324b 784 if ($htmlsafe) {
cb718de0 785 $ret .= htmlspecialchars($res[1]);
786 } else {
787 $ret .= $res[1];
788 }
098ea084 789 $encoding = ucfirst($res[3]);
d6f584fc 790
791 /* decide about valid decoding */
792 if ($decide && is_conversion_safe($res[2])) {
8bd0068d 793 $utfencode=true;
794 $can_be_encoded=true;
d6f584fc 795 } else {
8bd0068d 796 $can_be_encoded=false;
d6f584fc 797 }
098ea084 798 switch ($encoding)
799 {
8bd0068d 800 case 'B':
801 $replace = base64_decode($res[4]);
802 if ($utfencode) {
803 if ($can_be_encoded) {
804 /* convert string to different charset,
805 * if functions asks for it (usually in compose)
806 */
bcae324b 807 $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsafe);
8bd0068d 808 } else {
809 // convert string to html codes in order to display it
810 $ret .= charset_decode($res[2],$replace);
811 }
fab65ca9 812 } else {
bcae324b 813 if ($htmlsafe) {
8bd0068d 814 $replace = htmlspecialchars($replace);
815 }
816 $ret.= $replace;
fab65ca9 817 }
8bd0068d 818 break;
819 case 'Q':
820 $replace = str_replace('_', ' ', $res[4]);
821 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
822 $replace);
823 if ($utfencode) {
824 if ($can_be_encoded) {
825 /* convert string to different charset,
826 * if functions asks for it (usually in compose)
827 */
bcae324b 828 $replace = charset_convert($res[2], $replace,$default_charset,$htmlsafe);
8bd0068d 829 } else {
830 // convert string to html codes in order to display it
831 $replace = charset_decode($res[2], $replace);
832 }
833 } else {
bcae324b 834 if ($htmlsafe) {
8bd0068d 835 $replace = htmlspecialchars($replace);
836 }
098ea084 837 }
8bd0068d 838 $ret .= $replace;
839 break;
840 default:
841 break;
79e07c7e 842 }
098ea084 843 $chunk = $res[5];
844 $encoded = true;
08b7f7cc 845 }
846 if (!$encoded) {
bcae324b 847 if ($htmlsafe) {
9f7f68c3 848 $ret .= '&#32;';
08b7f7cc 849 } else {
850 $ret .= ' ';
da2415c1 851 }
08b7f7cc 852 }
dc3d13a7 853
bcae324b 854 if (!$encoded && $htmlsafe) {
dc3d13a7 855 $ret .= htmlspecialchars($chunk);
856 } else {
857 $ret .= $chunk;
858 }
098ea084 859 ++$i;
860 }
fd81e884 861 /* remove the first added space */
862 if ($ret) {
bcae324b 863 if ($htmlsafe) {
9f7f68c3 864 $ret = substr($ret,5);
fd81e884 865 } else {
866 $ret = substr($ret,1);
867 }
868 }
da2415c1 869
08b7f7cc 870 return $ret;
451f74a2 871}
872
9f7f68c3 873/**
a24cf710 874 * Encodes header
8bd0068d 875 *
a24cf710 876 * Function uses XTRA_CODE _encodeheader function, if such function exists.
a24cf710 877 *
758a7889 878 * Function uses Q encoding by default and encodes a string according to RFC
879 * 1522 for use in headers if it contains 8-bit characters or anything that
a24cf710 880 * looks like it should be encoded.
8bd0068d 881 *
758a7889 882 * Function switches to B encoding and encodeHeaderBase64() function, if
883 * string is 8bit and multibyte character set supported by mbstring extension
884 * is used. It can cause E_USER_NOTICE errors, if interface is used with
f270a6eb 885 * multibyte character set unsupported by mbstring extension.
886 *
8bd0068d 887 * @param string $string header string, that has to be encoded
888 * @return string quoted-printable encoded string
f270a6eb 889 * @todo make $mb_charsets system wide constant
8bd0068d 890 */
451f74a2 891function encodeHeader ($string) {
6fbd125b 892 global $default_charset, $languages, $squirrelmail_language;
83be314a 893
10dec454 894 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 895 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
33a55f5a 896 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
83be314a 897 }
793cc001 898
a24cf710 899 // Use B encoding for multibyte charsets
f270a6eb 900 $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
901 if (in_array($default_charset,$mb_charsets) &&
902 in_array($default_charset,sq_mb_list_encodings()) &&
903 sq_is8bit($string)) {
904 return encodeHeaderBase64($string,$default_charset);
905 } elseif (in_array($default_charset,$mb_charsets) &&
906 sq_is8bit($string) &&
907 ! in_array($default_charset,sq_mb_list_encodings())) {
908 // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
909 // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
910 }
a24cf710 911
451f74a2 912 // Encode only if the string contains 8-bit characters or =?
3d8371be 913 $j = strlen($string);
098ea084 914 $max_l = 75 - strlen($default_charset) - 7;
915 $aRet = array();
451f74a2 916 $ret = '';
c96c32f4 917 $iEncStart = $enc_init = false;
0d53f0f9 918 $cur_l = $iOffset = 0;
3d8371be 919 for($i = 0; $i < $j; ++$i) {
c96c32f4 920 switch($string{$i})
921 {
8bd0068d 922 case '=':
923 case '<':
924 case '>':
925 case ',':
926 case '?':
927 case '_':
928 if ($iEncStart === false) {
929 $iEncStart = $i;
930 }
931 $cur_l+=3;
932 if ($cur_l > ($max_l-2)) {
933 /* if there is an stringpart that doesn't need encoding, add it */
08b7f7cc 934 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
c96c32f4 935 $aRet[] = "=?$default_charset?Q?$ret?=";
936 $iOffset = $i;
937 $cur_l = 0;
938 $ret = '';
939 $iEncStart = false;
08b7f7cc 940 } else {
8bd0068d 941 $ret .= sprintf("=%02X",ord($string{$i}));
c96c32f4 942 }
8bd0068d 943 break;
944 case '(':
945 case ')':
946 if ($iEncStart !== false) {
08b7f7cc 947 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
8bd0068d 948 $aRet[] = "=?$default_charset?Q?$ret?=";
c96c32f4 949 $iOffset = $i;
8bd0068d 950 $cur_l = 0;
951 $ret = '';
952 $iEncStart = false;
c96c32f4 953 }
8bd0068d 954 break;
955 case ' ':
c96c32f4 956 if ($iEncStart !== false) {
098ea084 957 $cur_l++;
958 if ($cur_l > $max_l) {
08b7f7cc 959 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
c96c32f4 960 $aRet[] = "=?$default_charset?Q?$ret?=";
c96c32f4 961 $iOffset = $i;
962 $cur_l = 0;
098ea084 963 $ret = '';
8bd0068d 964 $iEncStart = false;
08b7f7cc 965 } else {
8bd0068d 966 $ret .= '_';
c96c32f4 967 }
3d8371be 968 }
8bd0068d 969 break;
970 default:
971 $k = ord($string{$i});
972 if ($k > 126) {
973 if ($iEncStart === false) {
974 // do not start encoding in the middle of a string, also take the rest of the word.
975 $sLeadString = substr($string,0,$i);
976 $aLeadString = explode(' ',$sLeadString);
977 $sToBeEncoded = array_pop($aLeadString);
978 $iEncStart = $i - strlen($sToBeEncoded);
979 $ret .= $sToBeEncoded;
980 $cur_l += strlen($sToBeEncoded);
981 }
982 $cur_l += 3;
983 /* first we add the encoded string that reached it's max size */
984 if ($cur_l > ($max_l-2)) {
985 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
986 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
987 $cur_l = 3;
988 $ret = '';
989 $iOffset = $i;
990 $iEncStart = $i;
991 }
992 $enc_init = true;
993 $ret .= sprintf("=%02X", $k);
994 } else {
995 if ($iEncStart !== false) {
996 $cur_l++;
997 if ($cur_l > $max_l) {
998 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
999 $aRet[] = "=?$default_charset?Q?$ret?=";
1000 $iEncStart = false;
1001 $iOffset = $i;
1002 $cur_l = 0;
1003 $ret = '';
1004 } else {
1005 $ret .= $string{$i};
1006 }
1007 }
1008 }
1009 break;
f7b3ba37 1010 }
451f74a2 1011 }
793cc001 1012
c96c32f4 1013 if ($enc_init) {
1014 if ($iEncStart !== false) {
1015 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1016 $aRet[] = "=?$default_charset?Q?$ret?=";
1017 } else {
1018 $aRet[] = substr($string,$iOffset);
1019 }
1020 $string = implode('',$aRet);
451f74a2 1021 }
3d8371be 1022 return $string;
451f74a2 1023}
b74ba498 1024
f270a6eb 1025/**
1026 * Encodes string according to rfc2047 B encoding header formating rules
1027 *
758a7889 1028 * It is recommended way to encode headers with character sets that store
f270a6eb 1029 * symbols in more than one byte.
1030 *
1031 * Function requires mbstring support. If required mbstring functions are missing,
1032 * function returns false and sets E_USER_WARNING level error message.
1033 *
758a7889 1034 * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
1035 * that mbstring functions will generate E_WARNING errors, if unsupported
f270a6eb 1036 * character set is used. mb_encode_mimeheader function provided by php
1037 * mbstring extension is not used in order to get better control of header
1038 * encoding.
1039 *
758a7889 1040 * Used php code functions - function_exists(), trigger_error(), strlen()
1041 * (is used with charset names and base64 strings). Used php mbstring
f270a6eb 1042 * functions - mb_strlen and mb_substr.
1043 *
758a7889 1044 * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
f270a6eb 1045 * encoding), rfc 2822 (header folding)
1046 *
1047 * @param string $string header string that must be encoded
758a7889 1048 * @param string $charset character set. Must be supported by mbstring extension.
f270a6eb 1049 * Use sq_mb_list_encodings() to detect supported charsets.
1050 * @return string string encoded according to rfc2047 B encoding formating rules
1051 * @since 1.5.1
1052 * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1053 * @todo Do we want to control max length of header?
1054 * @todo Do we want to control EOL (end-of-line) marker?
1055 * @todo Do we want to translate error message?
1056 */
1057function encodeHeaderBase64($string,$charset) {
1058 /**
1059 * Check mbstring function requirements.
1060 */
1061 if (! function_exists('mb_strlen') ||
1062 ! function_exists('mb_substr')) {
1063 // set E_USER_WARNING
1064 trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
1065 // return false
1066 return false;
1067 }
1068
1069 // initial return array
1070 $aRet = array();
1071
1072 /**
1073 * header length = 75 symbols max (same as in encodeHeader)
1074 * remove $charset length
1075 * remove =? ? ?= (5 chars)
1076 * remove 2 more chars (\r\n ?)
1077 */
1078 $iMaxLength = 75 - strlen($charset) - 7;
1079
1080 // set first character position
1081 $iStartCharNum = 0;
1082
1083 // loop through all characters. count characters and not bytes.
1084 for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1085 // encode string from starting character to current character.
1086 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1087
1088 // Check encoded string length
1089 if(strlen($encoded_string)>$iMaxLength) {
1090 // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1091 $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1092
1093 // set new starting character
1094 $iStartCharNum = $iCharNum-1;
1095
1096 // encode last char (in case it is last character in string)
1097 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1098 } // if string is shorter than max length - add next character
1099 }
1100
1101 // add last encoded string to array
1102 $aRet[] = $encoded_string;
1103
1104 // set initial return string
1105 $sRet = '';
1106
1107 // loop through encoded strings
1108 foreach($aRet as $string) {
1109 // TODO: Do we want to control EOL (end-of-line) marker
1110 if ($sRet!='') $sRet.= " ";
1111
1112 // add header tags and encoded string to return string
1113 $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1114 }
1115
1116 return $sRet;
1117}
1118
691a2d25 1119/* This function trys to locate the entity_id of a specific mime element */
3d8371be 1120function find_ent_id($id, $message) {
a171b359 1121 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1122 if ($message->entities[$i]->header->type0 == 'multipart') {
3d8371be 1123 $ret = find_ent_id($id, $message->entities[$i]);
451f74a2 1124 } else {
3d8371be 1125 if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
d8cffbab 1126// if (sq_check_save_extension($message->entities[$i])) {
8bd0068d 1127 return $message->entities[$i]->entity_id;
da2415c1 1128// }
c8f5f606 1129 } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1130 /**
1131 * This is part of a fix for Outlook Express 6.x generating
1132 * cid URLs without creating content-id headers
1133 * @@JA - 20050207
1134 */
1135 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1136 return $message->entities[$i]->entity_id;
1137 }
3d8371be 1138 }
a3daaaf3 1139 }
451f74a2 1140 }
3d8371be 1141 return $ret;
451f74a2 1142}
a3daaaf3 1143
e5e9381a 1144function sq_check_save_extension($message) {
1145 $filename = $message->getFilename();
1146 $ext = substr($filename, strrpos($filename,'.')+1);
1147 $save_extensions = array('jpg','jpeg','gif','png','bmp');
3d8371be 1148 return in_array($ext, $save_extensions);
e5e9381a 1149}
1150
1151
691a2d25 1152/**
8bd0068d 1153 ** HTMLFILTER ROUTINES
1154 */
451f74a2 1155
691a2d25 1156/**
0493ed11 1157 * This function checks attribute values for entity-encoded values
1158 * and returns them translated into 8-bit strings so we can run
1159 * checks on them.
8bd0068d 1160 *
0493ed11 1161 * @param $attvalue A string to run entity check against.
1162 * @return Nothing, modifies a reference value.
8bd0068d 1163 */
0493ed11 1164function sq_defang(&$attvalue){
1165 $me = 'sq_defang';
2dd879b8 1166 /**
0493ed11 1167 * Skip this if there aren't ampersands or backslashes.
8bd0068d 1168 */
0493ed11 1169 if (strpos($attvalue, '&') === false
1170 && strpos($attvalue, '\\') === false){
1171 return;
2dd879b8 1172 }
0493ed11 1173 $m = false;
2b665f28 1174 // before deent, translate the dangerous unicode characters and ... to safe values
1175 // otherwise the regular expressions do not match.
1176
1177
1178
0493ed11 1179 do {
1180 $m = false;
1181 $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1182 $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1183 $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1184 } while ($m == true);
1185 $attvalue = stripslashes($attvalue);
2dd879b8 1186}
1187
1188/**
8bd0068d 1189 * Kill any tabs, newlines, or carriage returns. Our friends the
1190 * makers of the browser with 95% market value decided that it'd
1191 * be funny to make "java[tab]script" be just as good as "javascript".
1192 *
1193 * @param attvalue The attribute value before extraneous spaces removed.
0493ed11 1194 * @return attvalue Nothing, modifies a reference value.
8bd0068d 1195 */
0493ed11 1196function sq_unspace(&$attvalue){
1197 $me = 'sq_unspace';
1198 if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1199 $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1200 Array('', '', '', '', ''), $attvalue);
2dd879b8 1201 }
2dd879b8 1202}
1203
1204/**
88de4926 1205 * Translate all dangerous Unicode or Shift_JIS characters which are accepted by
2b665f28 1206 * IE as regular characters.
1207 *
1208 * @param attvalue The attribute value before dangerous characters are translated.
1209 * @return attvalue Nothing, modifies a reference value.
1210 * @author Marc Groot Koerkamp.
1211 */
1212function sq_fixIE_idiocy(&$attvalue) {
1213 // remove NUL
1214 $attvalue = str_replace("\0", "", $attvalue);
1215 // remove comments
1216 $attvalue = preg_replace("/(\/\*.*?\*\/)/","",$attvalue);
1217
88de4926 1218 // IE has the evil habit of accepting every possible value for the attribute expression.
1219 // The table below contains characters which are parsed by IE if they are used in the "expression"
2b665f28 1220 // attribute value.
1221 $aDangerousCharsReplacementTable = array(
1222 array('&#x029F;', '&#0671;' ,/* L UNICODE IPA Extension */
1223 '&#x0280;', '&#0640;' ,/* R UNICODE IPA Extension */
1224 '&#x0274;', '&#0628;' ,/* N UNICODE IPA Extension */
567dc524 1225 '&#xFF25;', '&#65317;' ,/* Unicode FULLWIDTH LATIN CAPITAL LETTER E */
1226 '&#xFF45;', '&#65349;' ,/* Unicode FULLWIDTH LATIN SMALL LETTER E */
2b665f28 1227 '&#xFF38;', '&#65336;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER X */
1228 '&#xFF58;', '&#65368;',/* Unicode FULLWIDTH LATIN SMALL LETTER X */
1229 '&#xFF30;', '&#65328;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER P */
1230 '&#xFF50;', '&#65360;',/* Unicode FULLWIDTH LATIN SMALL LETTER P */
1231 '&#xFF32;', '&#65330;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER R */
1232 '&#xFF52;', '&#65362;',/* Unicode FULLWIDTH LATIN SMALL LETTER R */
1233 '&#xFF33;', '&#65331;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER S */
1234 '&#xFF53;', '&#65363;',/* Unicode FULLWIDTH LATIN SMALL LETTER S */
1235 '&#xFF29;', '&#65321;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER I */
1236 '&#xFF49;', '&#65353;',/* Unicode FULLWIDTH LATIN SMALL LETTER I */
1237 '&#xFF2F;', '&#65327;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER O */
1238 '&#xFF4F;', '&#65359;',/* Unicode FULLWIDTH LATIN SMALL LETTER O */
1239 '&#xFF2E;', '&#65326;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER N */
1240 '&#xFF4E;', '&#65358;',/* Unicode FULLWIDTH LATIN SMALL LETTER N */
1241 '&#xFF2C;', '&#65324;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER L */
1242 '&#xFF4C;', '&#65356;',/* Unicode FULLWIDTH LATIN SMALL LETTER L */
1243 '&#xFF35;', '&#65333;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER U */
1244 '&#xFF55;', '&#65365;',/* Unicode FULLWIDTH LATIN SMALL LETTER U */
1245 '&#x207F;', '&#8319;' ,/* Unicode SUPERSCRIPT LATIN SMALL LETTER N */
567dc524 1246 "\xEF\xBC\xA5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER E */ // in unicode this is some Chinese char range
1247 "\xEF\xBD\x85", /* Shift JIS FULLWIDTH LATIN SMALL LETTER E */
1248 "\xEF\xBC\xB8", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER X */
1249 "\xEF\xBD\x98", /* Shift JIS FULLWIDTH LATIN SMALL LETTER X */
1250 "\xEF\xBC\xB0", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER P */
1251 "\xEF\xBD\x90", /* Shift JIS FULLWIDTH LATIN SMALL LETTER P */
1252 "\xEF\xBC\xB2", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER R */
1253 "\xEF\xBD\x92", /* Shift JIS FULLWIDTH LATIN SMALL LETTER R */
1254 "\xEF\xBC\xB3", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER S */
1255 "\xEF\xBD\x93", /* Shift JIS FULLWIDTH LATIN SMALL LETTER S */
1256 "\xEF\xBC\xA9", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER I */
1257 "\xEF\xBD\x89", /* Shift JIS FULLWIDTH LATIN SMALL LETTER I */
1258 "\xEF\xBC\xAF", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER O */
1259 "\xEF\xBD\x8F", /* Shift JIS FULLWIDTH LATIN SMALL LETTER O */
1260 "\xEF\xBC\xAE", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER N */
1261 "\xEF\xBD\x8E", /* Shift JIS FULLWIDTH LATIN SMALL LETTER N */
1262 "\xEF\xBC\xAC", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER L */
1263 "\xEF\xBD\x8C", /* Shift JIS FULLWIDTH LATIN SMALL LETTER L */
1264 "\xEF\xBC\xB5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER U */
1265 "\xEF\xBD\x95", /* Shift JIS FULLWIDTH LATIN SMALL LETTER U */
1266 "\xE2\x81\xBF", /* Shift JIS FULLWIDTH SUPERSCRIPT N */
1267 "\xCA\x9F", /* L UNICODE IPA Extension */
1268 "\xCA\x80", /* R UNICODE IPA Extension */
1269 "\xC9\xB4"), /* N UNICODE IPA Extension */
2b665f28 1270 array('l', 'l', 'r','r','n','n',
567dc524 1271 'E','E','e','e','X','X','x','x','P','P','p','p','R','R','r','r','S','S','s','s','I','I',
1272 'i','i','O','O','o','o','N','N','n','n','L','L','l','l','U','U','u','u','n','n',
1273 '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 1274 $attvalue = str_replace($aDangerousCharsReplacementTable[0],$aDangerousCharsReplacementTable[1],$attvalue);
1275
88de4926 1276 // Escapes are useful for special characters like "{}[]()'&. In other cases they are
1277 // used for XSS.
2b665f28 1278 $attvalue = preg_replace("/(\\\\)([a-zA-Z]{1})/",'$2',$attvalue);
1279}
1280
1281/**
8bd0068d 1282 * This function returns the final tag out of the tag name, an array
1283 * of attributes, and the type of the tag. This function is called by
1284 * sq_sanitize internally.
1285 *
1286 * @param $tagname the name of the tag.
1287 * @param $attary the array of attributes and their values
1288 * @param $tagtype The type of the tag (see in comments).
1289 * @return a string with the final tag representation.
1290 */
691a2d25 1291function sq_tagprint($tagname, $attary, $tagtype){
b583c3e8 1292 $me = 'sq_tagprint';
3d8371be 1293
691a2d25 1294 if ($tagtype == 2){
1295 $fulltag = '</' . $tagname . '>';
1296 } else {
1297 $fulltag = '<' . $tagname;
1298 if (is_array($attary) && sizeof($attary)){
1299 $atts = Array();
1300 while (list($attname, $attvalue) = each($attary)){
1301 array_push($atts, "$attname=$attvalue");
1302 }
1303 $fulltag .= ' ' . join(" ", $atts);
1304 }
1305 if ($tagtype == 3){
b583c3e8 1306 $fulltag .= ' /';
691a2d25 1307 }
b583c3e8 1308 $fulltag .= '>';
451f74a2 1309 }
691a2d25 1310 return $fulltag;
451f74a2 1311}
a3daaaf3 1312
691a2d25 1313/**
8bd0068d 1314 * A small helper function to use with array_walk. Modifies a by-ref
1315 * value and makes it lowercase.
1316 *
1317 * @param $val a value passed by-ref.
1318 * @return void since it modifies a by-ref value.
1319 */
691a2d25 1320function sq_casenormalize(&$val){
1321 $val = strtolower($val);
1322}
451f74a2 1323
691a2d25 1324/**
8bd0068d 1325 * This function skips any whitespace from the current position within
1326 * a string and to the next non-whitespace value.
1327 *
1328 * @param $body the string
1329 * @param $offset the offset within the string where we should start
1330 * looking for the next non-whitespace character.
1331 * @return the location within the $body where the next
1332 * non-whitespace char is located.
1333 */
691a2d25 1334function sq_skipspace($body, $offset){
b583c3e8 1335 $me = 'sq_skipspace';
3d8371be 1336 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
691a2d25 1337 if (sizeof($matches{1})){
1338 $count = strlen($matches{1});
1339 $offset += $count;
451f74a2 1340 }
691a2d25 1341 return $offset;
451f74a2 1342}
a3daaaf3 1343
691a2d25 1344/**
8bd0068d 1345 * This function looks for the next character within a string. It's
1346 * really just a glorified "strpos", except it catches if failures
1347 * nicely.
1348 *
1349 * @param $body The string to look for needle in.
1350 * @param $offset Start looking from this position.
1351 * @param $needle The character/string to look for.
1352 * @return location of the next occurance of the needle, or
1353 * strlen($body) if needle wasn't found.
1354 */
691a2d25 1355function sq_findnxstr($body, $offset, $needle){
3d8371be 1356 $me = 'sq_findnxstr';
691a2d25 1357 $pos = strpos($body, $needle, $offset);
1358 if ($pos === FALSE){
1359 $pos = strlen($body);
451f74a2 1360 }
691a2d25 1361 return $pos;
451f74a2 1362}
a3daaaf3 1363
691a2d25 1364/**
8bd0068d 1365 * This function takes a PCRE-style regexp and tries to match it
1366 * within the string.
1367 *
1368 * @param $body The string to look for needle in.
1369 * @param $offset Start looking from here.
1370 * @param $reg A PCRE-style regex to match.
1371 * @return Returns a false if no matches found, or an array
1372 * with the following members:
1373 * - integer with the location of the match within $body
1374 * - string with whatever content between offset and the match
1375 * - string with whatever it is we matched
1376 */
691a2d25 1377function sq_findnxreg($body, $offset, $reg){
b583c3e8 1378 $me = 'sq_findnxreg';
691a2d25 1379 $matches = Array();
1380 $retarr = Array();
7d06541f 1381 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1382 if (!isset($matches{0}) || !$matches{0}){
691a2d25 1383 $retarr = false;
1384 } else {
1385 $retarr{0} = $offset + strlen($matches{1});
1386 $retarr{1} = $matches{1};
1387 $retarr{2} = $matches{2};
1388 }
1389 return $retarr;
1390}
a3daaaf3 1391
691a2d25 1392/**
8bd0068d 1393 * This function looks for the next tag.
1394 *
1395 * @param $body String where to look for the next tag.
1396 * @param $offset Start looking from here.
1397 * @return false if no more tags exist in the body, or
1398 * an array with the following members:
1399 * - string with the name of the tag
1400 * - array with attributes and their values
1401 * - integer with tag type (1, 2, or 3)
1402 * - integer where the tag starts (starting "<")
1403 * - integer where the tag ends (ending ">")
1404 * first three members will be false, if the tag is invalid.
1405 */
691a2d25 1406function sq_getnxtag($body, $offset){
b583c3e8 1407 $me = 'sq_getnxtag';
691a2d25 1408 if ($offset > strlen($body)){
1409 return false;
1410 }
1411 $lt = sq_findnxstr($body, $offset, "<");
1412 if ($lt == strlen($body)){
1413 return false;
1414 }
1415 /**
8bd0068d 1416 * We are here:
1417 * blah blah <tag attribute="value">
1418 * \---------^
1419 */
691a2d25 1420 $pos = sq_skipspace($body, $lt+1);
1421 if ($pos >= strlen($body)){
1422 return Array(false, false, false, $lt, strlen($body));
1423 }
1424 /**
8bd0068d 1425 * There are 3 kinds of tags:
1426 * 1. Opening tag, e.g.:
1427 * <a href="blah">
1428 * 2. Closing tag, e.g.:
1429 * </a>
1430 * 3. XHTML-style content-less tag, e.g.:
1431 * <img src="blah" />
1432 */
691a2d25 1433 $tagtype = false;
1434 switch (substr($body, $pos, 1)){
3d8371be 1435 case '/':
1436 $tagtype = 2;
1437 $pos++;
1438 break;
1439 case '!':
1440 /**
8bd0068d 1441 * A comment or an SGML declaration.
1442 */
3d8371be 1443 if (substr($body, $pos+1, 2) == "--"){
1444 $gt = strpos($body, "-->", $pos);
1445 if ($gt === false){
1446 $gt = strlen($body);
1447 } else {
1448 $gt += 2;
1449 }
1450 return Array(false, false, false, $lt, $gt);
bb8d0799 1451 } else {
3d8371be 1452 $gt = sq_findnxstr($body, $pos, ">");
1453 return Array(false, false, false, $lt, $gt);
1454 }
1455 break;
1456 default:
1457 /**
8bd0068d 1458 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1459 * later.
1460 */
3d8371be 1461 $tagtype = 1;
1462 break;
691a2d25 1463 }
a3daaaf3 1464
0493ed11 1465 $tag_start = $pos;
691a2d25 1466 $tagname = '';
1467 /**
8bd0068d 1468 * Look for next [\W-_], which will indicate the end of the tag name.
1469 */
691a2d25 1470 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1471 if ($regary == false){
1472 return Array(false, false, false, $lt, strlen($body));
1473 }
1474 list($pos, $tagname, $match) = $regary;
1475 $tagname = strtolower($tagname);
1476
1477 /**
8bd0068d 1478 * $match can be either of these:
1479 * '>' indicating the end of the tag entirely.
1480 * '\s' indicating the end of the tag name.
1481 * '/' indicating that this is type-3 xhtml tag.
1482 *
1483 * Whatever else we find there indicates an invalid tag.
1484 */
691a2d25 1485 switch ($match){
3d8371be 1486 case '/':
691a2d25 1487 /**
8bd0068d 1488 * This is an xhtml-style tag with a closing / at the
1489 * end, like so: <img src="blah" />. Check if it's followed
1490 * by the closing bracket. If not, then this tag is invalid
1491 */
3d8371be 1492 if (substr($body, $pos, 2) == "/>"){
1493 $pos++;
1494 $tagtype = 3;
1495 } else {
1496 $gt = sq_findnxstr($body, $pos, ">");
1497 $retary = Array(false, false, false, $lt, $gt);
1498 return $retary;
1499 }
1500 case '>':
1501 return Array($tagname, false, $tagtype, $lt, $pos);
1502 break;
1503 default:
1504 /**
8bd0068d 1505 * Check if it's whitespace
1506 */
3d8371be 1507 if (!preg_match('/\s/', $match)){
1508 /**
8bd0068d 1509 * This is an invalid tag! Look for the next closing ">".
1510 */
7d06541f 1511 $gt = sq_findnxstr($body, $lt, ">");
3d8371be 1512 return Array(false, false, false, $lt, $gt);
1513 }
1514 break;
691a2d25 1515 }
3d8371be 1516
691a2d25 1517 /**
8bd0068d 1518 * At this point we're here:
1519 * <tagname attribute='blah'>
1520 * \-------^
1521 *
1522 * At this point we loop in order to find all attributes.
1523 */
691a2d25 1524 $attname = '';
0493ed11 1525 $atttype = false;
691a2d25 1526 $attary = Array();
1527
1528 while ($pos <= strlen($body)){
1529 $pos = sq_skipspace($body, $pos);
1530 if ($pos == strlen($body)){
1531 /**
8bd0068d 1532 * Non-closed tag.
1533 */
691a2d25 1534 return Array(false, false, false, $lt, $pos);
1535 }
1536 /**
8bd0068d 1537 * See if we arrived at a ">" or "/>", which means that we reached
1538 * the end of the tag.
1539 */
691a2d25 1540 $matches = Array();
164800ad 1541 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
c828931c 1542 /**
8bd0068d 1543 * Yep. So we did.
1544 */
c828931c 1545 $pos += strlen($matches{1});
1546 if ($matches{2} == "/>"){
1547 $tagtype = 3;
1548 $pos++;
1549 }
1550 return Array($tagname, $attary, $tagtype, $lt, $pos);
1551 }
a3daaaf3 1552
cca46357 1553 /**
8bd0068d 1554 * There are several types of attributes, with optional
1555 * [:space:] between members.
1556 * Type 1:
1557 * attrname[:space:]=[:space:]'CDATA'
1558 * Type 2:
1559 * attrname[:space:]=[:space:]"CDATA"
1560 * Type 3:
1561 * attr[:space:]=[:space:]CDATA
1562 * Type 4:
1563 * attrname
1564 *
1565 * We leave types 1 and 2 the same, type 3 we check for
1566 * '"' and convert to "&quot" if needed, then wrap in
1567 * double quotes. Type 4 we convert into:
1568 * attrname="yes".
1569 */
3f7c623f 1570 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
691a2d25 1571 if ($regary == false){
1572 /**
8bd0068d 1573 * Looks like body ended before the end of tag.
1574 */
691a2d25 1575 return Array(false, false, false, $lt, strlen($body));
cca46357 1576 }
691a2d25 1577 list($pos, $attname, $match) = $regary;
1578 $attname = strtolower($attname);
1579 /**
8bd0068d 1580 * We arrived at the end of attribute name. Several things possible
1581 * here:
1582 * '>' means the end of the tag and this is attribute type 4
1583 * '/' if followed by '>' means the same thing as above
1584 * '\s' means a lot of things -- look what it's followed by.
1585 * anything else means the attribute is invalid.
1586 */
691a2d25 1587 switch($match){
3d8371be 1588 case '/':
691a2d25 1589 /**
8bd0068d 1590 * This is an xhtml-style tag with a closing / at the
1591 * end, like so: <img src="blah" />. Check if it's followed
1592 * by the closing bracket. If not, then this tag is invalid
1593 */
3d8371be 1594 if (substr($body, $pos, 2) == "/>"){
691a2d25 1595 $pos++;
3d8371be 1596 $tagtype = 3;
691a2d25 1597 } else {
3d8371be 1598 $gt = sq_findnxstr($body, $pos, ">");
1599 $retary = Array(false, false, false, $lt, $gt);
1600 return $retary;
1601 }
1602 case '>':
1603 $attary{$attname} = '"yes"';
1604 return Array($tagname, $attary, $tagtype, $lt, $pos);
1605 break;
1606 default:
1607 /**
8bd0068d 1608 * Skip whitespace and see what we arrive at.
1609 */
3d8371be 1610 $pos = sq_skipspace($body, $pos);
1611 $char = substr($body, $pos, 1);
1612 /**
8bd0068d 1613 * Two things are valid here:
1614 * '=' means this is attribute type 1 2 or 3.
1615 * \w means this was attribute type 4.
1616 * anything else we ignore and re-loop. End of tag and
1617 * invalid stuff will be caught by our checks at the beginning
1618 * of the loop.
1619 */
3d8371be 1620 if ($char == "="){
1621 $pos++;
1622 $pos = sq_skipspace($body, $pos);
691a2d25 1623 /**
8bd0068d 1624 * Here are 3 possibilities:
1625 * "'" attribute type 1
1626 * '"' attribute type 2
1627 * everything else is the content of tag type 3
1628 */
3d8371be 1629 $quot = substr($body, $pos, 1);
1630 if ($quot == "'"){
1631 $regary = sq_findnxreg($body, $pos+1, "\'");
1632 if ($regary == false){
1633 return Array(false, false, false, $lt, strlen($body));
1634 }
1635 list($pos, $attval, $match) = $regary;
1636 $pos++;
1637 $attary{$attname} = "'" . $attval . "'";
1638 } else if ($quot == '"'){
1639 $regary = sq_findnxreg($body, $pos+1, '\"');
1640 if ($regary == false){
1641 return Array(false, false, false, $lt, strlen($body));
1642 }
1643 list($pos, $attval, $match) = $regary;
1644 $pos++;
1645 $attary{$attname} = '"' . $attval . '"';
1646 } else {
1647 /**
8bd0068d 1648 * These are hateful. Look for \s, or >.
1649 */
3d8371be 1650 $regary = sq_findnxreg($body, $pos, "[\s>]");
1651 if ($regary == false){
1652 return Array(false, false, false, $lt, strlen($body));
1653 }
1654 list($pos, $attval, $match) = $regary;
1655 /**
8bd0068d 1656 * If it's ">" it will be caught at the top.
1657 */
3d8371be 1658 $attval = preg_replace("/\"/s", "&quot;", $attval);
1659 $attary{$attname} = '"' . $attval . '"';
7e235a1a 1660 }
3d8371be 1661 } else if (preg_match("|[\w/>]|", $char)) {
691a2d25 1662 /**
8bd0068d 1663 * That was attribute type 4.
1664 */
3d8371be 1665 $attary{$attname} = '"yes"';
1666 } else {
1667 /**
8bd0068d 1668 * An illegal character. Find next '>' and return.
1669 */
3d8371be 1670 $gt = sq_findnxstr($body, $pos, ">");
1671 return Array(false, false, false, $lt, $gt);
451f74a2 1672 }
3d8371be 1673 break;
691a2d25 1674 }
1675 }
1676 /**
8bd0068d 1677 * The fact that we got here indicates that the tag end was never
1678 * found. Return invalid tag indication so it gets stripped.
1679 */
691a2d25 1680 return Array(false, false, false, $lt, strlen($body));
1681}
1682
1683/**
0493ed11 1684 * Translates entities into literal values so they can be checked.
8bd0068d 1685 *
0493ed11 1686 * @param $attvalue the by-ref value to check.
1687 * @param $regex the regular expression to check against.
1688 * @param $hex whether the entites are hexadecimal.
1689 * @return True or False depending on whether there were matches.
8bd0068d 1690 */
0493ed11 1691function sq_deent(&$attvalue, $regex, $hex=false){
b583c3e8 1692 $me = 'sq_deent';
0493ed11 1693 $ret_match = false;
2b665f28 1694 // remove comments
1695 //$attvalue = preg_replace("/(\/\*.*\*\/)/","",$attvalue);
0493ed11 1696 preg_match_all($regex, $attvalue, $matches);
1697 if (is_array($matches) && sizeof($matches[0]) > 0){
1698 $repl = Array();
1699 for ($i = 0; $i < sizeof($matches[0]); $i++){
1700 $numval = $matches[1][$i];
1701 if ($hex){
1702 $numval = hexdec($numval);
a3daaaf3 1703 }
0493ed11 1704 $repl{$matches[0][$i]} = chr($numval);
691a2d25 1705 }
0493ed11 1706 $attvalue = strtr($attvalue, $repl);
1707 return true;
1708 } else {
1709 return false;
691a2d25 1710 }
691a2d25 1711}
1712
1713/**
8bd0068d 1714 * This function runs various checks against the attributes.
1715 *
1716 * @param $tagname String with the name of the tag.
1717 * @param $attary Array with all tag attributes.
1718 * @param $rm_attnames See description for sq_sanitize
1719 * @param $bad_attvals See description for sq_sanitize
1720 * @param $add_attr_to_tag See description for sq_sanitize
1721 * @param $message message object
1722 * @param $id message id
1723 * @return Array with modified attributes.
1724 */
da2415c1 1725function sq_fixatts($tagname,
1726 $attary,
691a2d25 1727 $rm_attnames,
1728 $bad_attvals,
1729 $add_attr_to_tag,
1730 $message,
b3af12ef 1731 $id,
3d8371be 1732 $mailbox
691a2d25 1733 ){
b583c3e8 1734 $me = 'sq_fixatts';
691a2d25 1735 while (list($attname, $attvalue) = each($attary)){
1736 /**
8bd0068d 1737 * See if this attribute should be removed.
1738 */
691a2d25 1739 foreach ($rm_attnames as $matchtag=>$matchattrs){
1740 if (preg_match($matchtag, $tagname)){
1741 foreach ($matchattrs as $matchattr){
1742 if (preg_match($matchattr, $attname)){
1743 unset($attary{$attname});
1744 continue;
1745 }
451f74a2 1746 }
451f74a2 1747 }
691a2d25 1748 }
1749 /**
2b665f28 1750 * Workaround for IE quirks
1751 */
1752 sq_fixIE_idiocy($attvalue);
1753
1754 /**
8bd0068d 1755 * Remove any backslashes, entities, and extraneous whitespace.
1756 */
2b665f28 1757
1758 $oldattvalue = $attvalue;
0493ed11 1759 sq_defang($attvalue);
2b665f28 1760 if ($attname == 'style' && $attvalue !== $oldattvalue) {
1761 // entities are used in the attribute value. In 99% of the cases it's there as XSS
1762 // i.e.<div style="{ left:exp&#x0280;essio&#x0274;( alert('XSS') ) }">
1763 $attvalue = "idiocy";
1764 $attary{$attname} = $attvalue;
1765 }
0493ed11 1766 sq_unspace($attvalue);
af861a34 1767
1768 /**
8bd0068d 1769 * Now let's run checks on the attvalues.
1770 * I don't expect anyone to comprehend this. If you do,
1771 * get in touch with me so I can drive to where you live and
1772 * shake your hand personally. :)
1773 */
691a2d25 1774 foreach ($bad_attvals as $matchtag=>$matchattrs){
1775 if (preg_match($matchtag, $tagname)){
1776 foreach ($matchattrs as $matchattr=>$valary){
1777 if (preg_match($matchattr, $attname)){
1778 /**
8bd0068d 1779 * There are two arrays in valary.
1780 * First is matches.
1781 * Second one is replacements
1782 */
691a2d25 1783 list($valmatch, $valrepl) = $valary;
da2415c1 1784 $newvalue =
691a2d25 1785 preg_replace($valmatch, $valrepl, $attvalue);
1786 if ($newvalue != $attvalue){
1787 $attary{$attname} = $newvalue;
567dc524 1788 $attvalue = $newvalue;
691a2d25 1789 }
1790 }
1791 }
451f74a2 1792 }
a3daaaf3 1793 }
567dc524 1794 if ($attname == 'style') {
1795 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1796 // 8bit and control characters in style attribute values can be used for XSS, remove them
1797 $attary{$attname} = '"disallowed character"';
1798 }
1799 preg_match_all("/url\s*\((.+)\)/si",$attvalue,$aMatch);
1800 if (count($aMatch)) {
1801 foreach($aMatch[1] as $sMatch) {
1802 // url value
1803 $urlvalue = $sMatch;
1804 sq_fix_url($attname, $urlvalue, $message, $id, $mailbox,"'");
1805 $attary{$attname} = str_replace($sMatch,$urlvalue,$attvalue);
1806 }
1807 }
691a2d25 1808 }
ff940ebc 1809 /**
567dc524 1810 * Use white list based filtering on attributes which can contain url's
ff940ebc 1811 */
567dc524 1812 else if ($attname == 'href' || $attname == 'src' || $attname == 'background') {
1813 sq_fix_url($attname, $attvalue, $message, $id, $mailbox);
1814 $attary{$attname} = $attvalue;
ff940ebc 1815 }
a3daaaf3 1816 }
691a2d25 1817 /**
8bd0068d 1818 * See if we need to append any attributes to this tag.
1819 */
691a2d25 1820 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1821 if (preg_match($matchtag, $tagname)){
1822 $attary = array_merge($attary, $addattary);
1823 }
1824 }
1825 return $attary;
451f74a2 1826}
a3daaaf3 1827
691a2d25 1828/**
567dc524 1829 * This function filters url's
1830 *
1831 * @param $attvalue String with attribute value to filter
1832 * @param $message message object
1833 * @param $id message id
1834 * @param $mailbox mailbox
1835 * @param $sQuote quoting characters around url's
1836 */
1837function sq_fix_url($attname, &$attvalue, $message, $id, $mailbox,$sQuote = '"') {
1838 $attvalue = trim($attvalue);
1839 if ($attvalue && ($attvalue[0] =='"'|| $attvalue[0] == "'")) {
1840 // remove the double quotes
1841 $sQuote = $attvalue[0];
1842 $attvalue = trim(substr($attvalue,1,-1));
1843 }
1844
d31f73f1 1845 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
1846 // images off by default.
afbf184c 1847 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET, FALSE);
d31f73f1 1848
567dc524 1849 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1850
1851 /**
1852 * Replace empty src tags with the blank image. src is only used
1853 * for frames, images, and image inputs. Doing a replace should
1854 * not affect them working as should be, however it will stop
1855 * IE from being kicked off when src for img tags are not set
1856 */
1857 if ($attvalue == '') {
1858 $attvalue = '"' . SM_PATH . 'images/blank.png"';
1859 } else {
1860 // first, disallow 8 bit characters and control characters
1861 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1862 switch ($attname) {
1863 case 'href':
1864 $attvalue = $sQuote . 'http://invalid-stuff-detected.example.com' . $sQuote;
1865 break;
1866 default:
1867 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1868 break;
1869 }
1870 } else {
1871 $aUrl = parse_url($attvalue);
1872 if (isset($aUrl['scheme'])) {
1873 switch(strtolower($aUrl['scheme'])) {
d75e755b 1874 case 'mailto':
567dc524 1875 case 'http':
1876 case 'https':
1877 case 'ftp':
1878 if ($attname != 'href') {
1879 if ($view_unsafe_images == false) {
1880 $attvalue = $sQuote . $secremoveimg . $sQuote;
1881 } else {
1882 if (isset($aUrl['path'])) {
8bbbc757 1883
1884 // No one has been able to show that image URIs
1885 // can be exploited, so for now, no restrictions
1886 // are made at all. If this proves to be a problem,
1887 // the commented-out code below can be of help.
1888 // (One consideration is that I see nothing in this
1889 // function that specifically says that we will
1890 // only ever arrive here when inspecting an image
1891 // tag, although that does seem to be the end
1892 // result - e.g., <script src="..."> where malicious
1893 // image URIs are in fact a problem are already
1894 // filtered out elsewhere.
1895 /* ---------------------------------
567dc524 1896 // validate image extension.
1897 $ext = strtolower(substr($aUrl['path'],strrpos($aUrl['path'],'.')));
1898 if (!in_array($ext,array('.jpeg','.jpg','xjpeg','.gif','.bmp','.jpe','.png','.xbm'))) {
8bbbc757 1899 // If URI is to something other than
1900 // a regular image file, get the contents
1901 // and try to see if it is an image.
1902 // Don't use Fileinfo (finfo_file()) because
1903 // we'd need to make the admin configure the
1904 // location of the magic.mime file (FIXME: add finfo_file() support later?)
1905 //
1906 $mime_type = '';
1907 if (function_exists('mime_content_type')
1908 && ($FILE = @fopen($attvalue, 'rb', FALSE))) {
1909
1910 // fetch file
1911 //
1912 $file_contents = '';
1913 while (!feof($FILE)) {
1914 $file_contents .= fread($FILE, 8192);
1915 }
1916 fclose($FILE);
1917
1918 // store file locally
1919 //
1920 global $attachment_dir, $username;
1921 $hashed_attachment_dir = getHashedDir($username, $attachment_dir);
1922 $localfilename = GenerateRandomString(32, '', 7);
1923 $full_localfilename = "$hashed_attachment_dir/$localfilename";
1924 while (file_exists($full_localfilename)) {
1925 $localfilename = GenerateRandomString(32, '', 7);
1926 $full_localfilename = "$hashed_attachment_dir/$localfilename";
1927 }
1928 $FILE = fopen("$hashed_attachment_dir/$localfilename", 'wb');
1929 fwrite($FILE, $file_contents);
1930 fclose($FILE);
1931
1932 // get mime type and remove file
1933 //
1934 $mime_type = mime_content_type("$hashed_attachment_dir/$localfilename");
1935 unlink("$hashed_attachment_dir/$localfilename");
1936 }
1937 // debug: echo "$attvalue FILE TYPE IS $mime_type<HR>";
1938 if (substr(strtolower($mime_type), 0, 5) != 'image') {
1939 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1940 }
567dc524 1941 }
8bbbc757 1942 --------------------------------- */
567dc524 1943 } else {
1944 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1945 }
1946 }
cf088f55 1947 } else {
1948 $attvalue = $sQuote . $attvalue . $sQuote;
567dc524 1949 }
1950 break;
1951 case 'outbind':
1952 /**
1953 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1954 * One day MS might actually make it match something useful, for now, falling
1955 * back to using cid2http, so we can grab the blank.png.
1956 */
cf088f55 1957 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
567dc524 1958 break;
1959 case 'cid':
1960 /**
1961 * Turn cid: urls into http-friendly ones.
1962 */
cf088f55 1963 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
567dc524 1964 break;
1965 default:
1966 $attvalue = $sQuote . SM_PATH . 'images/blank.png' . $sQuote;
1967 break;
1968 }
1969 } else {
1970 if (!(isset($aUrl['path']) && $aUrl['path'] == $secremoveimg)) {
1971 // parse_url did not lead to satisfying result
1972 $attvalue = $sQuote . SM_PATH . 'images/blank.png' . $sQuote;
1973 }
1974 }
1975 }
1976 }
1977}
1978
1979/**
8bd0068d 1980 * This function edits the style definition to make them friendly and
1981 * usable in SquirrelMail.
1982 *
1983 * @param $message the message object
1984 * @param $id the message id
1985 * @param $content a string with whatever is between <style> and </style>
1986 * @param $mailbox the message mailbox
1987 * @return a string with edited content.
1988 */
e60a299a 1989function sq_fixstyle($body, $pos, $message, $id, $mailbox){
b583c3e8 1990 $me = 'sq_fixstyle';
7e2ff844 1991 // workaround for </style> in between comments
1992 $iCurrentPos = $pos;
1993 $content = '';
1994 $sToken = '';
1995 $bSucces = false;
1996 $bEndTag = false;
1997 for ($i=$pos,$iCount=strlen($body);$i<$iCount;++$i) {
1998 $char = $body{$i};
1999 switch ($char) {
2000 case '<':
10c0caeb 2001 $sToken = $char;
7e2ff844 2002 break;
2003 case '/':
2004 if ($sToken == '<') {
2005 $sToken .= $char;
2006 $bEndTag = true;
2007 } else {
2008 $content .= $char;
2009 }
2010 break;
2011 case '>':
2012 if ($bEndTag) {
2013 $sToken .= $char;
2014 if (preg_match('/\<\/\s*style\s*\>/i',$sToken,$aMatch)) {
2015 $newpos = $i + 1;
2016 $bSucces = true;
2017 break 2;
2018 } else {
2019 $content .= $sToken;
2020 }
2021 $bEndTag = false;
2022 } else {
2023 $content .= $char;
2024 }
2025 break;
2026 case '!':
2027 if ($sToken == '<') {
2028 // possible comment
2029 if (isset($body{$i+2}) && substr($body,$i,3) == '!--') {
2030 $i = strpos($body,'-->',$i+3);
2b665f28 2031 if ($i === false) { // no end comment
2032 $i = strlen($body);
2033 }
7e2ff844 2034 $sToken = '';
2035 }
2036 } else {
2037 $content .= $char;
2038 }
2039 break;
2040 default:
2041 if ($bEndTag) {
2042 $sToken .= $char;
2043 } else {
2044 $content .= $char;
2045 }
2046 break;
2047 }
2048 }
2049 if ($bSucces == FALSE){
7d06541f 2050 return array(FALSE, strlen($body));
2051 }
7e2ff844 2052
2b665f28 2053
2054
691a2d25 2055 /**
9ebace19 2056 * First look for general BODY style declaration, which would be
2057 * like so:
2058 * body {background: blah-blah}
2059 * and change it to .bodyclass so we can just assign it to a <div>
2060 */
691a2d25 2061 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
3d8371be 2062 $secremoveimg = '../images/' . _("sec_remove_eng.png");
691a2d25 2063 /**
0493ed11 2064 * Fix url('blah') declarations.
2065 */
2066 // $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
2067 // "url(\\1$secremoveimg\\2)", $content);
2b665f28 2068
567dc524 2069 // first check for 8bit sequences and disallowed control characters
2070 if (preg_match('/[\16-\37\200-\377]+/',$content)) {
2071 $content = '<!-- style block removed by html filter due to presence of 8bit characters -->';
2072 return array($content, $newpos);
2073 }
2074
2b665f28 2075 // IE Sucks hard. We have a special function for it.
2076 sq_fixIE_idiocy($content);
2077
2078 // remove @import line
2079 $content = preg_replace("/^\s*(@import.*)$/mi","\n<!-- @import rules forbidden -->\n",$content);
2080
5b4884be 2081 // translate ur\l and variations (IE parses that)
2b665f28 2082 // TODO check if the sq_fixIE_idiocy function already handles this.
5b4884be 2083 $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
567dc524 2084 preg_match_all("/url\s*\((.+)\)/si",$content,$aMatch);
2085 if (count($aMatch)) {
2086 $aValue = $aReplace = array();
2087 foreach($aMatch[1] as $sMatch) {
2088 // url value
2089 $urlvalue = $sMatch;
2090 sq_fix_url('style',$urlvalue, $message, $id, $mailbox,"'");
2091 $aValue[] = $sMatch;
2092 $aReplace[] = $urlvalue;
0493ed11 2093 }
567dc524 2094 $content = str_replace($aValue,$aReplace,$content);
691a2d25 2095 }
567dc524 2096
9ebace19 2097 /**
2098 * Remove any backslashes, entities, and extraneous whitespace.
2099 */
0493ed11 2100 $contentTemp = $content;
2101 sq_defang($contentTemp);
2102 sq_unspace($contentTemp);
a3daaaf3 2103
691a2d25 2104 /**
8bd0068d 2105 * Fix stupid css declarations which lead to vulnerabilities
2106 * in IE.
2107 */
5db90261 2108 $match = Array('/\/\*.*\*\//',
2109 '/expression/i',
0493ed11 2110 '/behaviou*r/i',
2111 '/binding/i',
2b665f28 2112 '/include-source/i',
2113 '/javascript/i',
2114 '/script/i');
2115 $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy');
0493ed11 2116 $contentNew = preg_replace($match, $replace, $contentTemp);
2117 if ($contentNew !== $contentTemp) {
2118 // insecure css declarations are used. From now on we don't care
2119 // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
2120 $content = $contentNew;
2121 }
7d06541f 2122 return array($content, $newpos);
691a2d25 2123}
a3daaaf3 2124
0493ed11 2125
691a2d25 2126/**
8bd0068d 2127 * This function converts cid: url's into the ones that can be viewed in
2128 * the browser.
2129 *
2130 * @param $message the message object
2131 * @param $id the message id
2132 * @param $cidurl the cid: url.
2133 * @param $mailbox the message mailbox
2134 * @return a string with a http-friendly url
2135 */
b3af12ef 2136function sq_cid2http($message, $id, $cidurl, $mailbox){
691a2d25 2137 /**
8bd0068d 2138 * Get rid of quotes.
2139 */
691a2d25 2140 $quotchar = substr($cidurl, 0, 1);
2dd879b8 2141 if ($quotchar == '"' || $quotchar == "'"){
2142 $cidurl = str_replace($quotchar, "", $cidurl);
2143 } else {
2144 $quotchar = '';
2145 }
691a2d25 2146 $cidurl = substr(trim($cidurl), 4);
0493ed11 2147
2148 $match_str = '/\{.*?\}\//';
2149 $str_rep = '';
2150 $cidurl = preg_replace($match_str, $str_rep, $cidurl);
2151
e5e9381a 2152 $linkurl = find_ent_id($cidurl, $message);
9ebace19 2153 /* in case of non-safe cid links $httpurl should be replaced by a sort of
2154 unsafe link image */
e5e9381a 2155 $httpurl = '';
c8f5f606 2156
8bd0068d 2157 /**
2158 * This is part of a fix for Outlook Express 6.x generating
2159 * cid URLs without creating content-id headers. These images are
2160 * not part of the multipart/related html mail. The html contains
2161 * <img src="cid:{some_id}/image_filename.ext"> references to
2162 * attached images with as goal to render them inline although
2163 * the attachment disposition property is not inline.
2164 */
c8f5f606 2165
2166 if (empty($linkurl)) {
2167 if (preg_match('/{.*}\//', $cidurl)) {
2168 $cidurl = preg_replace('/{.*}\//','', $cidurl);
2169 if (!empty($cidurl)) {
2170 $linkurl = find_ent_id($cidurl, $message);
2171 }
2172 }
2173 }
f8a1ed5a 2174
c8f5f606 2175 if (!empty($linkurl)) {
6b04287c 2176 $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
8bd0068d 2177 "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
2178 '&amp;ent_id=' . $linkurl . $quotchar;
c8f5f606 2179 } else {
2180 /**
2181 * If we couldn't generate a proper img url, drop in a blank image
2182 * instead of sending back empty, otherwise it causes unusual behaviour
2183 */
bc017c1d 2184 $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
e5e9381a 2185 }
f8a1ed5a 2186
691a2d25 2187 return $httpurl;
2188}
2189
2190/**
8bd0068d 2191 * This function changes the <body> tag into a <div> tag since we
2192 * can't really have a body-within-body.
2193 *
2194 * @param $attary an array of attributes and values of <body>
2195 * @param $mailbox mailbox we're currently reading (for cid2http)
2196 * @param $message current message (for cid2http)
2197 * @param $id current message id (for cid2http)
2198 * @return a modified array of attributes to be set for <div>
2199 */
2dd879b8 2200function sq_body2div($attary, $mailbox, $message, $id){
b583c3e8 2201 $me = 'sq_body2div';
3d8371be 2202 $divattary = Array('class' => "'bodyclass'");
b583c3e8 2203 $text = '#000000';
c189a963 2204 $has_bgc_stl = $has_txt_stl = false;
b583c3e8 2205 $styledef = '';
691a2d25 2206 if (is_array($attary) && sizeof($attary) > 0){
2207 foreach ($attary as $attname=>$attvalue){
2208 $quotchar = substr($attvalue, 0, 1);
2209 $attvalue = str_replace($quotchar, "", $attvalue);
2210 switch ($attname){
3d8371be 2211 case 'background':
8bd0068d 2212 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
3d8371be 2213 $styledef .= "background-image: url('$attvalue'); ";
2214 break;
2215 case 'bgcolor':
c189a963 2216 $has_bgc_stl = true;
3d8371be 2217 $styledef .= "background-color: $attvalue; ";
2218 break;
2219 case 'text':
c189a963 2220 $has_txt_stl = true;
3d8371be 2221 $styledef .= "color: $attvalue; ";
2222 break;
691a2d25 2223 }
a3daaaf3 2224 }
c189a963 2225 // Outlook defines a white bgcolor and no text color. This can lead to
2226 // white text on a white bg with certain themes.
2227 if ($has_bgc_stl && !$has_txt_stl) {
2228 $styledef .= "color: $text; ";
2229 }
691a2d25 2230 if (strlen($styledef) > 0){
2231 $divattary{"style"} = "\"$styledef\"";
2232 }
2233 }
2234 return $divattary;
2235}
a3daaaf3 2236
691a2d25 2237/**
8bd0068d 2238 * This is the main function and the one you should actually be calling.
2239 * There are several variables you should be aware of an which need
2240 * special description.
2241 *
2242 * Since the description is quite lengthy, see it here:
2243 * http://linux.duke.edu/projects/mini/htmlfilter/
2244 *
2245 * @param $body the string with HTML you wish to filter
2246 * @param $tag_list see description above
2247 * @param $rm_tags_with_content see description above
2248 * @param $self_closing_tags see description above
2249 * @param $force_tag_closing see description above
2250 * @param $rm_attnames see description above
2251 * @param $bad_attvals see description above
2252 * @param $add_attr_to_tag see description above
2253 * @param $message message object
2254 * @param $id message id
2255 * @return sanitized html safe to show on your pages.
2256 */
da2415c1 2257function sq_sanitize($body,
8bd0068d 2258 $tag_list,
2259 $rm_tags_with_content,
2260 $self_closing_tags,
2261 $force_tag_closing,
2262 $rm_attnames,
2263 $bad_attvals,
2264 $add_attr_to_tag,
2265 $message,
2266 $id,
2267 $mailbox
2268 ){
b583c3e8 2269 $me = 'sq_sanitize';
7d06541f 2270 $rm_tags = array_shift($tag_list);
691a2d25 2271 /**
8bd0068d 2272 * Normalize rm_tags and rm_tags_with_content.
2273 */
7d06541f 2274 @array_walk($tag_list, 'sq_casenormalize');
691a2d25 2275 @array_walk($rm_tags_with_content, 'sq_casenormalize');
2276 @array_walk($self_closing_tags, 'sq_casenormalize');
2277 /**
8bd0068d 2278 * See if tag_list is of tags to remove or tags to allow.
2279 * false means remove these tags
2280 * true means allow these tags
2281 */
691a2d25 2282 $curpos = 0;
2283 $open_tags = Array();
2dd879b8 2284 $trusted = "\n<!-- begin sanitized html -->\n";
691a2d25 2285 $skip_content = false;
bb8d0799 2286 /**
8bd0068d 2287 * Take care of netscape's stupid javascript entities like
2288 * &{alert('boo')};
2289 */
bb8d0799 2290 $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
691a2d25 2291
7d06541f 2292 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
691a2d25 2293 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
2294 $free_content = substr($body, $curpos, $lt-$curpos);
2295 /**
8bd0068d 2296 * Take care of <style>
2297 */
7d06541f 2298 if ($tagname == "style" && $tagtype == 1){
da2415c1 2299 list($free_content, $curpos) =
e60a299a 2300 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
7d06541f 2301 if ($free_content != FALSE){
2302 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2303 $trusted .= $free_content;
2304 $trusted .= sq_tagprint($tagname, false, 2);
2305 }
2306 continue;
691a2d25 2307 }
2308 if ($skip_content == false){
2309 $trusted .= $free_content;
691a2d25 2310 }
2311 if ($tagname != FALSE){
2312 if ($tagtype == 2){
2313 if ($skip_content == $tagname){
2314 /**
8bd0068d 2315 * Got to the end of tag we needed to remove.
2316 */
691a2d25 2317 $tagname = false;
2318 $skip_content = false;
2319 } else {
2320 if ($skip_content == false){
c828931c 2321 if ($tagname == "body"){
2322 $tagname = "div";
2dd879b8 2323 }
da2415c1 2324 if (isset($open_tags{$tagname}) &&
8bd0068d 2325 $open_tags{$tagname} > 0){
2dd879b8 2326 $open_tags{$tagname}--;
691a2d25 2327 } else {
2dd879b8 2328 $tagname = false;
691a2d25 2329 }
691a2d25 2330 }
2331 }
2332 } else {
2333 /**
8bd0068d 2334 * $rm_tags_with_content
2335 */
691a2d25 2336 if ($skip_content == false){
2337 /**
8bd0068d 2338 * See if this is a self-closing type and change
2339 * tagtype appropriately.
2340 */
691a2d25 2341 if ($tagtype == 1
8bd0068d 2342 && in_array($tagname, $self_closing_tags)){
2dd879b8 2343 $tagtype = 3;
691a2d25 2344 }
2345 /**
8bd0068d 2346 * See if we should skip this tag and any content
2347 * inside it.
2348 */
691a2d25 2349 if ($tagtype == 1 &&
8bd0068d 2350 in_array($tagname, $rm_tags_with_content)){
691a2d25 2351 $skip_content = $tagname;
2352 } else {
da2415c1 2353 if (($rm_tags == false
8bd0068d 2354 && in_array($tagname, $tag_list)) ||
2355 ($rm_tags == true &&
2356 !in_array($tagname, $tag_list))){
691a2d25 2357 $tagname = false;
2358 } else {
2dd879b8 2359 /**
8bd0068d 2360 * Convert body into div.
2361 */
2dd879b8 2362 if ($tagname == "body"){
2363 $tagname = "div";
da2415c1 2364 $attary = sq_body2div($attary, $mailbox,
8bd0068d 2365 $message, $id);
2dd879b8 2366 }
691a2d25 2367 if ($tagtype == 1){
2368 if (isset($open_tags{$tagname})){
2369 $open_tags{$tagname}++;
2370 } else {
2371 $open_tags{$tagname}=1;
2372 }
2373 }
2374 /**
8bd0068d 2375 * This is where we run other checks.
2376 */
691a2d25 2377 if (is_array($attary) && sizeof($attary) > 0){
2378 $attary = sq_fixatts($tagname,
8bd0068d 2379 $attary,
2380 $rm_attnames,
2381 $bad_attvals,
2382 $add_attr_to_tag,
2383 $message,
2384 $id,
2385 $mailbox
2386 );
691a2d25 2387 }
2388 }
2389 }
691a2d25 2390 }
2391 }
2392 if ($tagname != false && $skip_content == false){
2393 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2394 }
691a2d25 2395 }
2396 $curpos = $gt+1;
a3daaaf3 2397 }
691a2d25 2398 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2399 if ($force_tag_closing == true){
2400 foreach ($open_tags as $tagname=>$opentimes){
2401 while ($opentimes > 0){
2402 $trusted .= '</' . $tagname . '>';
2403 $opentimes--;
2404 }
2405 }
2406 $trusted .= "\n";
2407 }
2408 $trusted .= "<!-- end sanitized html -->\n";
2409 return $trusted;
2410}
451f74a2 2411
691a2d25 2412/**
8bd0068d 2413 * This is a wrapper function to call html sanitizing routines.
2414 *
2415 * @param $body the body of the message
2416 * @param $id the id of the message
c189a963 2417
2418 * @param $message
2419 * @param $mailbox
2420 * @param boolean $take_mailto_links When TRUE, converts mailto: links
2421 * into internal SM compose links
2422 * (optional; default = TRUE)
8bd0068d 2423 * @return a string with html safe to display in the browser.
2424 */
c189a963 2425function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2426
202bcbcc 2427 // require_once(SM_PATH . 'functions/url_parser.php'); // for $MailTo_PReg_Match
c189a963 2428
691a2d25 2429 global $attachment_common_show_images, $view_unsafe_images,
8bd0068d 2430 $has_unsafe_images;
691a2d25 2431 /**
8bd0068d 2432 * Don't display attached images in HTML mode.
2b665f28 2433 *
d0187bd6 2434 * SB: why?
8bd0068d 2435 */
691a2d25 2436 $attachment_common_show_images = false;
2437 $tag_list = Array(
8bd0068d 2438 false,
2439 "object",
2440 "meta",
2441 "html",
2442 "head",
2443 "base",
2444 "link",
2445 "frame",
2446 "iframe",
2447 "plaintext",
2448 "marquee"
2449 );
691a2d25 2450
2451 $rm_tags_with_content = Array(
8bd0068d 2452 "script",
2453 "applet",
2454 "embed",
2455 "title",
2456 "frameset",
0493ed11 2457 "xmp",
8bd0068d 2458 "xml"
2459 );
691a2d25 2460
2461 $self_closing_tags = Array(
8bd0068d 2462 "img",
2463 "br",
2464 "hr",
2465 "input",
2466 "outbind"
2467 );
691a2d25 2468
2dd879b8 2469 $force_tag_closing = true;
691a2d25 2470
2471 $rm_attnames = Array(
8bd0068d 2472 "/.*/" =>
2473 Array(
2474 "/target/i",
2475 "/^on.*/i",
2476 "/^dynsrc/i",
2477 "/^data.*/i",
2478 "/^lowsrc.*/i"
2479 )
2480 );
691a2d25 2481
2482 $secremoveimg = "../images/" . _("sec_remove_eng.png");
2483 $bad_attvals = Array(
8bd0068d 2484 "/.*/" =>
691a2d25 2485 Array(
0a6ec9b5 2486 "/^src|background/i" =>
8bd0068d 2487 Array(
691a2d25 2488 Array(
8bd0068d 2489 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2490 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2491 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
691a2d25 2492 ),
8bd0068d 2493 Array(
2494 "\\1$secremoveimg\\2",
2495 "\\1$secremoveimg\\2",
2496 "\\1$secremoveimg\\2",
8bd0068d 2497 )
2498 ),
0a6ec9b5 2499 "/^href|action/i" =>
8bd0068d 2500 Array(
0a6ec9b5 2501 Array(
8bd0068d 2502 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2503 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2504 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
0a6ec9b5 2505 ),
691a2d25 2506 Array(
8bd0068d 2507 "\\1#\\1",
2508 "\\1#\\1",
8bd0068d 2509 "\\1#\\1"
02474e43 2510 )
8bd0068d 2511 ),
2512 "/^style/i" =>
2513 Array(
2514 Array(
5db90261 2515 "/\/\*.*\*\//",
8bd0068d 2516 "/expression/i",
2517 "/binding/i",
2518 "/behaviou*r/i",
2519 "/include-source/i",
2520 "/position\s*:\s*absolute/i",
1d935bc2 2521 "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
8bd0068d 2522 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2523 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2524 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2525 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2526 ),
2527 Array(
5db90261 2528 "",
8bd0068d 2529 "idiocy",
2530 "idiocy",
2531 "idiocy",
2532 "idiocy",
567dc524 2533 "idiocy",
5b4884be 2534 "url",
8bd0068d 2535 "url(\\1#\\1)",
2536 "url(\\1#\\1)",
2537 "url(\\1#\\1)",
8bd0068d 2538 "\\1:url(\\2#\\3)"
2539 )
691a2d25 2540 )
8bd0068d 2541 )
691a2d25 2542 );
9ebace19 2543
d31f73f1 2544 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
2545 // images off by default.
afbf184c 2546 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET, FALSE);
9ebace19 2547
691a2d25 2548 if (!$view_unsafe_images){
2549 /**
8bd0068d 2550 * Remove any references to http/https if view_unsafe_images set
2551 * to false.
2552 */
02474e43 2553 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
8bd0068d 2554 '/^([\'\"])\s*https*:.*([\'\"])/si');
02474e43 2555 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
8bd0068d 2556 "\\1$secremoveimg\\1");
02474e43 2557 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
7ef0b415 2558 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
02474e43 2559 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
8bd0068d 2560 "url(\\1$secremoveimg\\1)");
691a2d25 2561 }
451f74a2 2562
691a2d25 2563 $add_attr_to_tag = Array(
8bd0068d 2564 "/^a$/i" =>
c25f2fbb 2565 Array('target'=>'"_blank"',
02474e43 2566 'title'=>'"'._("This external link will open in a new window").'"'
8bd0068d 2567 )
2568 );
da2415c1 2569 $trusted = sq_sanitize($body,
8bd0068d 2570 $tag_list,
2571 $rm_tags_with_content,
2572 $self_closing_tags,
2573 $force_tag_closing,
2574 $rm_attnames,
2575 $bad_attvals,
2576 $add_attr_to_tag,
2577 $message,
2578 $id,
2579 $mailbox
2580 );
567dc524 2581 if (strpos($trusted,$secremoveimg)){
691a2d25 2582 $has_unsafe_images = true;
da2415c1 2583 }
c189a963 2584
2585 // we want to parse mailto's in HTML output, change to SM compose links
2586 // this is a modified version of code from url_parser.php... but Marc is
2587 // right: we need a better filtering implementation; adding this randomly
2588 // here is not a great solution
2589 //
2590 if ($take_mailto_links) {
2591 // parseUrl($trusted); // this even parses URLs inside of tags... too aggressive
2592 global $MailTo_PReg_Match;
202bcbcc 2593 $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1) ;
c189a963 2594 if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2595 foreach ($regs[0] as $i => $mailto_before) {
2596 $mailto_params = $regs[10][$i];
2597 // get rid of any tailing quote since we have to add send_to to the end
2598 //
2599 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2600 $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2601 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2602 $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2603
2604 if ($regs[1][$i]) { //if there is an email addr before '?', we need to merge it with the params
2605 $to = 'to=' . $regs[1][$i];
2606 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
2607 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2608 else {
2609 if ($mailto_params) //already some params, append to them
2610 $mailto_params .= '&amp;' . $to;
2611 else
2612 $mailto_params .= '?' . $to;
2613 }
2614 }
2615
2616 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2617
2618 // we'll already have target=_blank, no need to allow comp_in_new
2619 // here (which would be a lot more work anyway)
2620 //
2621 global $compose_new_win;
2622 $temp_comp_in_new = $compose_new_win;
2623 $compose_new_win = 0;
2624 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2625 $compose_new_win = $temp_comp_in_new;
2626
2627 // remove <a href=" and anything after the next quote (we only
2628 // need the uri, not the link HTML) in compose uri
2629 //
2630 $comp_uri = substr($comp_uri, 9);
2631 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2632 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2633 }
2634 }
2635 }
2636
691a2d25 2637 return $trusted;
451f74a2 2638}
a4a70693 2639
da2415c1 2640/**
8bd0068d 2641 * function SendDownloadHeaders - send file to the browser
2642 *
2643 * Original Source: SM core src/download.php
2644 * moved here to make it available to other code, and separate
2645 * front end from back end functionality.
2646 *
2647 * @param string $type0 first half of mime type
2648 * @param string $type1 second half of mime type
2649 * @param string $filename filename to tell the browser for downloaded file
2650 * @param boolean $force whether to force the download dialog to pop
2651 * @param optional integer $filesize send the Content-Header and length to the browser
2652 * @return void
2653 */
02474e43 2654function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2655 global $languages, $squirrelmail_language;
cfffd60b 2656 $isIE = $isIE6plus = false;
02474e43 2657
2658 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2659
2660 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
8bd0068d 2661 strstr($HTTP_USER_AGENT, 'Opera') === false) {
cfffd60b 2662 $isIE = true;
02474e43 2663 }
2664
cfffd60b 2665 if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2666 ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2667 $isIE6plus = true;
02474e43 2668 }
2669
2670 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 2671 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
02474e43 2672 $filename =
8bd0068d 2673 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
02474e43 2674 } else {
2675 $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
2676 }
2677
2678 // A Pox on Microsoft and it's Internet Explorer!
2679 //
2680 // IE has lots of bugs with file downloads.
2681 // It also has problems with SSL. Both of these cause problems
2682 // for us in this function.
2683 //
2684 // See this article on Cache Control headers and SSL
2685 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2686 //
2687 // The best thing you can do for IE is to upgrade to the latest
2688 // version
2689 //set all the Cache Control Headers for IE
2690 if ($isIE) {
2691 $filename=rawurlencode($filename);
2692 header ("Pragma: public");
8bd0068d 2693 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
02474e43 2694 header ("Cache-Control: post-check=0, pre-check=0", false);
8bd0068d 2695 header ("Cache-Control: private");
02474e43 2696
2697 //set the inline header for IE, we'll add the attachment header later if we need it
2698 header ("Content-Disposition: inline; filename=$filename");
2699 }
2700
2701 if (!$force) {
2702 // Try to show in browser window
2703 header ("Content-Disposition: inline; filename=\"$filename\"");
2704 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2705 } else {
2706 // Try to pop up the "save as" box
2707
2708 // IE makes this hard. It pops up 2 save boxes, or none.
2709 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2710 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2711 // But, according to Microsoft, it is "RFC compliant but doesn't
2712 // take into account some deviations that allowed within the
2713 // specification." Doesn't that mean RFC non-compliant?
2714 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2715
2716 // all browsers need the application/octet-stream header for this
2717 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2718
2719 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2720 // Do not have quotes around filename, but that applied to
2721 // "attachment"... does it apply to inline too?
2722 header ("Content-Disposition: attachment; filename=\"$filename\"");
2723
cfffd60b 2724 if ($isIE && !$isIE6plus) {
02474e43 2725 // This combination seems to work mostly. IE 5.5 SP 1 has
2726 // known issues (see the Microsoft Knowledge Base)
2727
2728 // This works for most types, but doesn't work with Word files
2729 header ("Content-Type: application/download; name=\"$filename\"");
7e2ff844 2730 header ("Content-Type: application/force-download; name=\"$filename\"");
02474e43 2731 // These are spares, just in case. :-)
2732 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2733 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2734 //header("Content-Type: application/octet-stream; name=\"$filename\"");
7e2ff844 2735 } else if ($isIE) {
2736 // This is to prevent IE for MIME sniffing and auto open a file in IE
2737 header ("Content-Type: application/force-download; name=\"$filename\"");
02474e43 2738 } else {
2739 // another application/octet-stream forces download for Netscape
2740 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2741 }
2742 }
2743
2744 //send the content-length header if the calling function provides it
2745 if ($filesize > 0) {
2746 header("Content-Length: $filesize");
2747 }
07c49f57 2748
8d863f64 2749} // end fn SendDownloadHeaders