Stop using session_unregister()
[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 *
d4e46166 9 * @copyright &copy; 1999-2009 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 $msg->is_flagged = true;
94 }
1a753239 95 else if (strtolower($flag) == '$forwarded') {
96 $msg->is_forwarded = true;
97 }
3d8371be 98 break;
99 case 'M':
100 if (strtolower($flag) == '$mdnsent') {
101 $msg->is_mdnsent = true;
102 }
103 break;
104 default:
105 break;
7a9e9c89 106 }
107 }
451f74a2 108 }
7a9e9c89 109 // listEntities($msg);
3d8371be 110 return $msg;
451f74a2 111}
b74ba498 112
22efa9fb 113
114
3d8371be 115/* This starts the parsing of a particular structure. It is called recursively,
8bd0068d 116 * so it can be passed different structures. It returns an object of type
117 * $message.
118 * First, it checks to see if it is a multipart message. If it is, then it
119 * handles that as it sees is necessary. If it is just a regular entity,
120 * then it parses it and adds the necessary header information (by calling out
121 * to mime_get_elements()
122 */
451f74a2 123
4d592352 124function mime_fetch_body($imap_stream, $id, $ent_id=1, $fetch_size=0) {
3d8371be 125 /* Do a bit of error correction. If we couldn't find the entity id, just guess
8bd0068d 126 * that it is the first one. That is usually the case anyway.
127 */
7c7b74b3 128
09a4bde3 129 if (!$ent_id) {
08b7f7cc 130 $cmd = "FETCH $id BODY[]";
1035e159 131 } else {
08b7f7cc 132 $cmd = "FETCH $id BODY[$ent_id]";
09a4bde3 133 }
3d8371be 134
4d592352 135 if ($fetch_size!=0) $cmd .= "<0.$fetch_size>";
da2415c1 136
6201339c 137 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, TRUE);
77b88425 138 do {
3d8371be 139 $topline = trim(array_shift($data));
140 } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
a4a70693 141
451f74a2 142 $wholemessage = implode('', $data);
143 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
3d8371be 144 $ret = substr($wholemessage, 0, $regs[1]);
145 /* There is some information in the content info header that could be important
8bd0068d 146 * in order to parse html messages. Let's get them here.
147 */
0600bdf1 148// if ($ret{0} == '<') {
6201339c 149// $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, TRUE);
0600bdf1 150// }
451f74a2 151 } else if (ereg('"([^"]*)"', $topline, $regs)) {
152 $ret = $regs[1];
06bcb9c3 153 } else if ((stristr($topline, 'nil') !== false) && (empty($wholemessage))) {
154 $ret = $wholemessage;
451f74a2 155 } else {
156 global $where, $what, $mailbox, $passed_id, $startMessage;
3d8371be 157 $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
451f74a2 158 if (isset($where) && isset($what)) {
3d8371be 159 $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
a3daaaf3 160 } else {
3d8371be 161 $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
451f74a2 162 }
e5ea9327 163 $par .= '&amp;response=' . urlencode($response) .
8bd0068d 164 '&amp;message=' . urlencode($message) .
165 '&amp;topline=' . urlencode($topline);
a019eeb8 166
8bd0068d 167 echo '<tt><br />' .
02474e43 168 '<table width="80%"><tr>' .
169 '<tr><td colspan="2">' .
170 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
171 '</td></tr>' .
172 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
173 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
174 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
175 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
176 "</table><br /></tt></font><hr />";
346817d4 177
6201339c 178 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, TRUE);
451f74a2 179 array_shift($data);
180 $wholemessage = implode('', $data);
a019eeb8 181
346817d4 182 $ret = $wholemessage;
a3daaaf3 183 }
3d8371be 184 return $ret;
451f74a2 185}
d4467150 186
fa26ab45 187function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding, $rStream='php://stdout', $force_crlf='') {
1035e159 188
3d8371be 189 /* Don't kill the connection if the browser is over a dialup
8bd0068d 190 * and it would take over 30 seconds to download it.
191 * Don't call set_time_limit in safe mode.
192 */
b7206e1d 193
3d8371be 194 if (!ini_get('safe_mode')) {
b7206e1d 195 set_time_limit(0);
196 }
7c7b74b3 197 /* in case of base64 encoded attachments, do not buffer them.
8bd0068d 198 Instead, echo the decoded attachment directly to screen */
7c7b74b3 199 if (strtolower($encoding) == 'base64') {
200 if (!$ent_id) {
7591f143 201 $query = "FETCH $id BODY[]";
7c7b74b3 202 } else {
7591f143 203 $query = "FETCH $id BODY[$ent_id]";
7c7b74b3 204 }
7591f143 205 sqimap_run_command($imap_stream,$query,true,$response,$message,TRUE,'sqimap_base64_decode',$rStream,true);
1d142b8d 206 } else {
7591f143 207 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
208 if (is_resource($rStream)) {
fa26ab45 209 fputs($rStream,decodeBody($body, $encoding, $force_crlf));
7591f143 210 } else {
fa26ab45 211 echo decodeBody($body, $encoding, $force_crlf);
7591f143 212 }
1d142b8d 213 }
346817d4 214
da2415c1 215 /*
8bd0068d 216 TODO, use the same method for quoted printable.
217 However, I assume that quoted printable attachments aren't that large
218 so the performancegain / memory usage drop will be minimal.
219 If we decide to add that then we need to adapt sqimap_fread because
220 we need to split te result on \n and fread doesn't stop at \n. That
221 means we also should provide $results from sqimap_fread (by ref) to
222 te function and set $no_return to false. The $filter function for
223 quoted printable should handle unsetting of $results.
224 */
da2415c1 225 /*
8bd0068d 226 TODO 2: find out how we write to the output stream php://stdout. fwrite
227 doesn't work because 'php://stdout isn't a stream.
228 */
7c7b74b3 229
5d9c6f73 230 return;
451f74a2 231}
beb9e459 232
451f74a2 233/* -[ END MIME DECODING ]----------------------------------------------------------- */
d4467150 234
3d8371be 235/* This is here for debugging purposes. It will print out a list
8bd0068d 236 * of all the entity IDs that are in the $message object.
237 */
451f74a2 238function listEntities ($message) {
3d8371be 239 if ($message) {
3c621ba1 240 echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br />';
3d8371be 241 for ($i = 0; isset($message->entities[$i]); $i++) {
242 echo "$i : ";
243 $msg = listEntities($message->entities[$i]);
244
245 if ($msg) {
246 echo "return: ";
247 return $msg;
248 }
249 }
a4a70693 250 }
451f74a2 251}
f0c4dc12 252
f792c641 253function getPriorityStr($priority) {
3d8371be 254 $priority_level = substr($priority,0,1);
255
256 switch($priority_level) {
257 /* Check for a higher then normal priority. */
258 case '1':
259 case '2':
260 $priority_string = _("High");
261 break;
262
263 /* Check for a lower then normal priority. */
264 case '4':
265 case '5':
266 $priority_string = _("Low");
267 break;
268
269 /* Check for a normal priority. */
270 case '3':
271 default:
272 $priority_level = '3';
273 $priority_string = _("Normal");
274 break;
275
276 }
277 return $priority_string;
f792c641 278}
279
451f74a2 280/* returns a $message object for a particular entity id */
281function getEntity ($message, $ent_id) {
a4a70693 282 return $message->getEntity($ent_id);
451f74a2 283}
8beafbbc 284
3d8371be 285/* translateText
8bd0068d 286 * Extracted from strings.php 23/03/2002
287 */
da4c66e8 288
289function translateText(&$body, $wrap_at, $charset) {
3d8371be 290 global $where, $what; /* from searching */
291 global $color; /* color theme */
da4c66e8 292
202bcbcc 293 // require_once(SM_PATH . 'functions/url_parser.php');
da4c66e8 294
295 $body_ary = explode("\n", $body);
da4c66e8 296 for ($i=0; $i < count($body_ary); $i++) {
4c1f4be3 297 $line = rtrim($body_ary[$i],"\r");
298
da4c66e8 299 if (strlen($line) - 2 >= $wrap_at) {
c7aff938 300 sqWordWrap($line, $wrap_at, $charset);
da4c66e8 301 }
302 $line = charset_decode($charset, $line);
303 $line = str_replace("\t", ' ', $line);
304
305 parseUrl ($line);
306
3d8371be 307 $quotes = 0;
da4c66e8 308 $pos = 0;
3d8371be 309 $j = strlen($line);
da4c66e8 310
3d8371be 311 while ($pos < $j) {
da4c66e8 312 if ($line[$pos] == ' ') {
3d8371be 313 $pos++;
da4c66e8 314 } else if (strpos($line, '&gt;', $pos) === $pos) {
315 $pos += 4;
3d8371be 316 $quotes++;
da4c66e8 317 } else {
318 break;
319 }
320 }
3d8371be 321
83c94382 322 if ($quotes % 2) {
d0814c02 323 $line = '<span class="quote1">' . $line . '</span>';
4c25967c 324 } elseif ($quotes) {
d0814c02 325 $line = '<span class="quote2">' . $line . '</span>';
da4c66e8 326 }
3d8371be 327
da4c66e8 328 $body_ary[$i] = $line;
329 }
330 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
331}
332
a2bfcbce 333/**
da1b55ad 334 * This returns a parsed string called $body. That string can then
335 * be displayed as the actual message in the HTML. It contains
336 * everything needed, including HTML Tags, Attachments at the
337 * bottom, etc.
f8a1ed5a 338 *
da1b55ad 339 * Since 1.2.0 function uses message_body hook.
340 * Till 1.3.0 function included output of formatAttachments().
341 *
342 * @param resource $imap_stream imap connection resource
343 * @param object $message squirrelmail message object
344 * @param array $color squirrelmail color theme array
345 * @param integer $wrap_at number of characters per line
346 * @param string $ent_num (since 1.3.0) message part id
347 * @param integer $id (since 1.3.0) message id
348 * @param string $mailbox (since 1.3.0) imap folder name
da1b55ad 349 * @return string html formated message text
350 */
a540f994 351function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
3d8371be 352 /* This if statement checks for the entity to show as the
8bd0068d 353 * primary message. To add more of them, just put them in the
354 * order that is their priority.
355 */
ce68b76b 356 global $startMessage, $languages, $squirrelmail_language,
40a34e57 357 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id,
14c85e39 358 $use_iframe, $iframe_height, $download_and_unsafe_link,
955bfc8f 359 $download_href, $unsafe_image_toggle_href, $unsafe_image_toggle_text,
551c7b53 360 $oTemplate, $nbsp;
2c25d36a 361
362 // workaround for not updated config.php
363 if (! isset($use_iframe)) $use_iframe = false;
77bfbd2e 364
d31f73f1 365 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
366 // images off by default.
afbf184c 367 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET, FALSE);
d03c24f4 368
cc34b00d 369 $body = '';
23bcec6f 370 $urlmailbox = urlencode($mailbox);
451f74a2 371 $body_message = getEntity($message, $ent_num);
372 if (($body_message->header->type0 == 'text') ||
8bd0068d 373 ($body_message->header->type0 == 'rfc822')) {
3d8371be 374 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
451f74a2 375 $body = decodeBody($body, $body_message->header->encoding);
e842b215 376
377 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 378 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
e842b215 379 if (mb_detect_encoding($body) != 'ASCII') {
33a55f5a 380 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
e842b215 381 }
382 }
d849b570 383
384 /* As of 1.5.2, $body is passed (and modified) by reference */
385 do_hook('message_body', $body);
23bcec6f 386
3d8371be 387 /* If there are other types that shouldn't be formatted, add
8bd0068d 388 * them here.
389 */
3d8371be 390
451f74a2 391 if ($body_message->header->type1 == 'html') {
3d8371be 392 if ($show_html_default <> 1) {
85015544 393 $entity_conv = array('&nbsp;' => ' ',
8bd0068d 394 '<p>' => "\n",
395 '<P>' => "\n",
396 '<br>' => "\n",
397 '<BR>' => "\n",
398 '<br />' => "\n",
399 '<BR />' => "\n",
400 '&gt;' => '>',
401 '&lt;' => '<');
85015544 402 $body = strtr($body, $entity_conv);
3d8371be 403 $body = strip_tags($body);
85015544 404 $body = trim($body);
405 translateText($body, $wrap_at,
8bd0068d 406 $body_message->header->getParameter('charset'));
a540f994 407 } elseif ($use_iframe) {
84410f31 408 /**
409 * If we don't add html message between iframe tags,
410 * we must detect unsafe images and modify $has_unsafe_images.
f8a1ed5a 411 */
758a7889 412 $html_body = magicHTML($body, $id, $message, $mailbox);
b6c52e61 413 // Convert character set in order to display html mails in different character set
414 $html_body = charset_decode($body_message->header->getParameter('charset'),$html_body,false,true);
84410f31 415
2c25d36a 416 // creating iframe url
417 $iframeurl=sqm_baseuri().'src/view_html.php?'
f8a1ed5a 418 . 'mailbox=' . $urlmailbox
2c25d36a 419 . '&amp;passed_id=' . $id
420 . '&amp;ent_id=' . $ent_num
421 . '&amp;view_unsafe_images=' . (int) $view_unsafe_images;
422
79d58d4c 423 global $oTemplate;
424 $oTemplate->assign('iframe_url', $iframeurl);
51554708 425 $oTemplate->assign('iframe_height', $iframe_height);
79d58d4c 426 $oTemplate->assign('html_body', $html_body);
2b665f28 427
79d58d4c 428 $body = $oTemplate->fetch('read_html_iframe.tpl');
a3daaaf3 429 } else {
2c25d36a 430 // old way of html rendering
b6c52e61 431 /**
758a7889 432 * convert character set. charset_decode does not remove html special chars
b6c52e61 433 * applied by magicHTML functions and does not sanitize them second time if
758a7889 434 * fourth argument is true.
435 */
567dc524 436 $charset = $body_message->header->getParameter('charset');
437 if (!empty($charset)) {
438 $body = charset_decode($charset,$body,false,true);
439 }
440 $body = magicHTML($body, $id, $message, $mailbox);
a3daaaf3 441 }
451f74a2 442 } else {
3d8371be 443 translateText($body, $wrap_at,
8bd0068d 444 $body_message->header->getParameter('charset'));
451f74a2 445 }
a2bfcbce 446
9ebace19 447 /*
448 * Previously the links for downloading and unsafe images were printed
449 * under the mail. By putting the links in a global variable we can
450 * print it in the toolbar where it belongs. Since the original code was
451 * in this place it's left here. It might be possible to move it to some
452 * other place if that makes sense. The possibility to do so has not
453 * been evaluated yet.
454 */
455
456 // Initialize the global variable to an empty string.
457 // 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 458 $download_and_unsafe_link = '';
459
9ebace19 460 // Prepare and build a link for downloading the mail.
83cf04bd 461 $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
8bd0068d 462 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
463 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
08b7f7cc 464 if (isset($passed_ent_id)) {
465 $link .= '&amp;passed_ent_id='.$passed_ent_id;
466 }
14c85e39 467 $download_href = SM_PATH . 'src/download.php?absolute_dl=true&amp;' . $link;
9ebace19 468
469 // Always add the link for downloading the mail as a file to the global
470 // variable.
955bfc8f 471 $download_and_unsafe_link .= "$nbsp|$nbsp"
472 . create_hyperlink($download_href, _("Download this as a file"));
9ebace19 473
474 // Find out the right text to use in the link depending on the
475 // circumstances. If the unsafe images are displayed the link should
476 // hide them, if they aren't displayed the link should only appear if
477 // the mail really contains unsafe images.
7aad7b77 478 if ($view_unsafe_images) {
23f617b8 479 $text = _("Hide Unsafe Images");
7aad7b77 480 } else {
08b7f7cc 481 if (isset($has_unsafe_images) && $has_unsafe_images) {
482 $link .= '&amp;view_unsafe_images=1';
483 $text = _("View Unsafe Images");
484 } else {
485 $text = '';
486 }
3d8371be 487 }
9ebace19 488
489 // Only create a link for unsafe images if there's need for one. If so:
490 // add it to the global variable.
83cf04bd 491 if($text != '') {
14c85e39 492 $unsafe_image_toggle_href = SM_PATH . 'src/read_body.php?'.$link;
493 $unsafe_image_toggle_text = $text;
955bfc8f 494 $download_and_unsafe_link .= "$nbsp|$nbsp"
495 . create_hyperlink($unsafe_image_toggle_href, $text);
83cf04bd 496 }
3d8371be 497 }
498 return $body;
451f74a2 499}
b74ba498 500
da1b55ad 501/**
a540f994 502 * Generate attachments array for passing to templates.
2b665f28 503 *
d67f519a 504 * @since 1.5.2
da1b55ad 505 * @param object $message SquirrelMail message object
506 * @param array $exclude_id message parts that are not attachments.
507 * @param string $mailbox mailbox name
508 * @param integer $id message id
da1b55ad 509 */
d67f519a 510function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
511 global $where, $what, $startMessage, $color, $passed_ent_id, $base_uri;
451f74a2 512
23bcec6f 513 $att_ar = $message->getAttachments($exclude_id);
23bcec6f 514 $urlMailbox = urlencode($mailbox);
515
d67f519a 516 $attachments = array();
23bcec6f 517 foreach ($att_ar as $att) {
fdc9d9b5 518 $ent = $att->entity_id;
2e25760a 519 $header = $att->header;
f0c4dc12 520 $type0 = strtolower($header->type0);
521 $type1 = strtolower($header->type1);
2e25760a 522 $name = '';
d0187bd6 523 $links = array();
21dab2dc 524 $links['download link']['text'] = _("Download");
202bcbcc 525 $links['download link']['href'] = $base_uri .
8bd0068d 526 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
2b665f28 527
2e25760a 528 if ($type0 =='message' && $type1 == 'rfc822') {
202bcbcc 529 $default_page = $base_uri . 'src/read_body.php';
2e25760a 530 $rfc822_header = $att->rfc822_header;
098ea084 531 $filename = $rfc822_header->subject;
6cc08d8b 532 if (trim( $filename ) == '') {
533 $filename = 'untitled-[' . $ent . ']' ;
08b7f7cc 534 }
2e25760a 535 $from_o = $rfc822_header->from;
536 if (is_object($from_o)) {
04ea844e 537 $from_name = decodeHeader($from_o->getAddress(false));
7e697748 538 } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
539 // something weird happens when a digest message is opened and you return to the digest
540 // now the from object is part of an array. Probably the parseHeader call overwrites the info
541 // retrieved from the bodystructure in a different way. We need to fix this later.
542 // possible starting point, do not fetch header we already have and inspect how
543 // the rfc822_header object behaves.
544 $from_name = decodeHeader($from_o[0]->getAddress(false));
2e25760a 545 } else {
546 $from_name = _("Unknown sender");
f0c4dc12 547 }
d0187bd6 548 $description = _("From").': '.$from_name;
23bcec6f 549 } else {
202bcbcc 550 $default_page = $base_uri . 'src/download.php';
02474e43 551 $filename = $att->getFilename();
f810c0b2 552 if ($header->description) {
098ea084 553 $description = decodeHeader($header->description);
f810c0b2 554 } else {
3d8371be 555 $description = '';
556 }
2e25760a 557 }
558
559 $display_filename = $filename;
3d8371be 560 if (isset($passed_ent_id)) {
561 $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
562 } else {
563 $passed_ent_id_link = '';
564 }
565 $defaultlink = $default_page . "?startMessage=$startMessage"
8bd0068d 566 . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
567 . '&amp;ent_id='.$ent.$passed_ent_id_link;
2e25760a 568 if ($where && $what) {
8bd0068d 569 $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
2e25760a 570 }
7e2ff844 571 // IE does make use of mime content sniffing. Forcing a download
572 // prohibit execution of XSS inside an application/octet-stream attachment
573 if ($type0 == 'application' && $type1 == 'octet-stream') {
574 $defaultlink .= '&amp;absolute_dl=true';
575 }
2b665f28 576
3d8371be 577 /* This executes the attachment hook with a specific MIME-type.
8bd0068d 578 * If that doesn't have results, it tries if there's a rule
9b94c54d 579 * for a more generic type. Finally, a hook for ALL attachment
580 * types is run as well.
8bd0068d 581 */
8dca4d22 582 // First remember the default link.
583 $defaultlink_orig = $defaultlink;
584
d849b570 585 /* The API for this hook has changed as of 1.5.2 so that all plugin
586 arguments are passed in an array instead of each their own plugin
587 argument, and arguments are passed by reference, so instead of
588 returning any changes, changes should simply be made to the original
589 arguments themselves. */
9c3b2d22 590 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
591 &$defaultlink, &$display_filename, &$where, &$what);
592 do_hook("attachment $type0/$type1", $temp);
8dca4d22 593 if(count($links) <= 1 && $defaultlink == $defaultlink_orig) {
d849b570 594 /* The API for this hook has changed as of 1.5.2 so that all plugin
595 arguments are passed in an array instead of each their own plugin
596 argument, and arguments are passed by reference, so instead of
597 returning any changes, changes should simply be made to the original
598 arguments themselves. */
9c3b2d22 599 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
600 &$defaultlink, &$display_filename, &$where, &$what);
601 do_hook("attachment $type0/*", $temp);
2e25760a 602 }
d849b570 603 /* The API for this hook has changed as of 1.5.2 so that all plugin
604 arguments are passed in an array instead of each their own plugin
605 argument, and arguments are passed by reference, so instead of
606 returning any changes, changes should simply be made to the original
607 arguments themselves. */
9c3b2d22 608 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
609 &$defaultlink, &$display_filename, &$where, &$what);
8dca4d22 610 // Do not let a generic plugin change the default link if a more
611 // specialized one already did it...
612 if ($defaultlink != $defaultlink_orig) {
613 $dummy = '';
614 $temp[5] = &$dummy;
615 }
9c3b2d22 616 do_hook("attachment */*", $temp);
77b88425 617
d67f519a 618 $this_attachment = array();
619 $this_attachment['Name'] = decodeHeader($display_filename);
620 $this_attachment['Description'] = $description;
621 $this_attachment['DefaultHREF'] = $defaultlink;
622 $this_attachment['DownloadHREF'] = $links['download link']['href'];
623 $this_attachment['ViewHREF'] = isset($links['attachment_common']) ? $links['attachment_common']['href'] : '';
624 $this_attachment['Size'] = $header->size;
625 $this_attachment['ContentType'] = htmlspecialchars($type0 .'/'. $type1);
626 $this_attachment['OtherLinks'] = array();
3d8371be 627 foreach ($links as $val) {
d0187bd6 628 if ($val['text']==_("Download") || $val['text'] == _("View"))
629 continue;
630 if (empty($val['text']) && empty($val['extra']))
631 continue;
2b665f28 632
d67f519a 633 $temp = array();
634 $temp['HREF'] = $val['href'];
635 $temp['Text'] = (empty($val['text']) ? '' : $val['text']) . (empty($val['extra']) ? '' : $val['extra']);
636 $this_attachment['OtherLinks'][] = $temp;
2e25760a 637 }
d67f519a 638 $attachments[] = $this_attachment;
2b665f28 639
3d8371be 640 unset($links);
2e25760a 641 }
2b665f28 642
d67f519a 643 return $attachments;
644}
645
646/**
647 * Displays attachment links and information
648 *
649 * Since 1.3.0 function is not included in formatBody() call.
650 *
651 * Since 1.0.2 uses attachment $type0/$type1 hook.
652 * Since 1.2.5 uses attachment $type0/* hook.
653 * Since 1.5.0 uses attachments_bottom hook.
654 * Since 1.5.2 uses templates and does *not* return a value.
655 *
656 * @param object $message SquirrelMail message object
657 * @param array $exclude_id message parts that are not attachments.
658 * @param string $mailbox mailbox name
659 * @param integer $id message id
660 */
661function formatAttachments($message, $exclude_id, $mailbox, $id) {
662 global $oTemplate;
2b665f28 663
d67f519a 664 $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
d0187bd6 665
666 $oTemplate->assign('attachments', $attach);
667 $oTemplate->display('read_attachments.tpl');
451f74a2 668}
b74ba498 669
7c7b74b3 670function sqimap_base64_decode(&$string) {
7c0ec1d8 671
b17a8968 672 // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
7c0ec1d8 673 // fly decoding (to reduce memory usage) you have to check if the
674 // data has incomplete pairs
675
b17a8968 676 // Remove the noise in order to check if the 4 bytes pairs are complete
7c0ec1d8 677 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
678
42ce44f8 679 $sStringRem = '';
7c0ec1d8 680 $iMod = strlen($string) % 4;
681 if ($iMod) {
682 $sStringRem = substr($string,-$iMod);
b17a8968 683 // Check if $sStringRem contains padding characters
7c0ec1d8 684 if (substr($sStringRem,-1) != '=') {
685 $string = substr($string,0,-$iMod);
686 } else {
687 $sStringRem = '';
688 }
689 }
7c7b74b3 690 $string = base64_decode($string);
7c0ec1d8 691 return $sStringRem;
7c7b74b3 692}
693
fdf7cef1 694/**
fa26ab45 695 * Decodes encoded string (usually message body)
696 *
697 * This function decodes a string (usually the message body)
698 * depending on the encoding type. Currently quoted-printable
699 * and base64 encodings are supported.
700 *
701 * The decode_body hook was added to this function in 1.4.2/1.5.0.
702 * The $force_crlf parameter was added in 1.5.2.
703 *
704 * @param string $string The encoded string
705 * @param string $encoding used encoding
706 * @param string $force_crlf Whether or not to force CRLF or LF
707 * line endings (or to leave as is).
708 * If given as "LF", line endings will
709 * all be converted to LF; if "CRLF",
710 * line endings will all be converted
711 * to CRLF. If given as an empty value,
9785376e 712 * the global $force_crlf_default will
fa26ab45 713 * be consulted (it can be specified in
714 * config/config_local.php). Otherwise,
715 * any other value will cause the string
716 * to be left alone. Note that this will
717 * be overridden to "LF" if not using at
718 * least PHP version 4.3.0. (OPTIONAL;
719 * default is empty - consult global
720 * default value)
721 *
722 * @return string The decoded string
fdf7cef1 723 *
fdf7cef1 724 * @since 1.0
fa26ab45 725 *
fdf7cef1 726 */
fa26ab45 727function decodeBody($string, $encoding, $force_crlf='') {
728
729 global $force_crlf_default;
730 if (empty($force_crlf)) $force_crlf = $force_crlf_default;
731 $force_crlf = strtoupper($force_crlf);
732
733 // must force line endings to LF due to broken
734 // quoted_printable_decode() in PHP versions
735 // before 4.3.0 (see below)
736 //
737 if (!check_php_version(4, 3, 0) || $force_crlf == 'LF')
738 $string = str_replace("\r\n", "\n", $string);
739 else if ($force_crlf == 'CRLF')
740 $string = str_replace("\n", "\r\n", $string);
83be314a 741
b583c3e8 742 $encoding = strtolower($encoding);
3d8371be 743
d849b570 744 $encoding_handler = do_hook('decode_body', $encoding);
5166f86a 745
746
fa26ab45 747 // plugins get first shot at decoding the string
5166f86a 748 //
749 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
fa26ab45 750 $string = $encoding_handler('decode', $string);
5166f86a 751
fdf7cef1 752 } elseif ($encoding == 'quoted-printable' ||
8bd0068d 753 $encoding == 'quoted_printable') {
fa26ab45 754
755 // quoted_printable_decode() function is broken in older
756 // php versions. Text with \r\n decoding was fixed only
757 // in php 4.3.0. Minimal code requirement is PHP 4.0.4+
758 // and the above call to: str_replace("\r\n", "\n", $string);
759 //
760 $string = quoted_printable_decode($string);
761
fdf7cef1 762 } elseif ($encoding == 'base64') {
fa26ab45 763 $string = base64_decode($string);
b583c3e8 764 }
3d8371be 765
b583c3e8 766 // All other encodings are returned raw.
fa26ab45 767 return $string;
451f74a2 768}
769
9f7f68c3 770/**
8bd0068d 771 * Decodes headers
772 *
e1115979 773 * This function decodes strings that are encoded according to
8bd0068d 774 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
775 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
776 *
777 * @param string $string header string that has to be made readable
778 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
bcae324b 779 * @param boolean $htmlsafe preserve spaces and sanitize html special characters. defaults to true
8bd0068d 780 * @param boolean $decide decide if string can be utfencoded. defaults to false
781 * @return string decoded header string
782 */
bcae324b 783function decodeHeader ($string, $utfencode=true,$htmlsafe=true,$decide=false) {
d6f584fc 784 global $languages, $squirrelmail_language,$default_charset;
79e07c7e 785 if (is_array($string)) {
786 $string = implode("\n", $string);
787 }
da2415c1 788
10dec454 789 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 790 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
33a55f5a 791 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
08b7f7cc 792 // Do we need to return at this point?
793 // return $string;
83be314a 794 }
79e07c7e 795 $i = 0;
08b7f7cc 796 $iLastMatch = -2;
db65b6b0 797 $encoded = true;
0a06275a 798
098ea084 799 $aString = explode(' ',$string);
08b7f7cc 800 $ret = '';
098ea084 801 foreach ($aString as $chunk) {
358a78a1 802 if ($encoded && $chunk === '') {
08b7f7cc 803 continue;
358a78a1 804 } elseif ($chunk === '') {
08b7f7cc 805 $ret .= ' ';
806 continue;
807 }
098ea084 808 $encoded = false;
08b7f7cc 809 /* if encoded words are not separated by a linear-space-white we still catch them */
810 $j = $i-1;
7e6ca3e8 811
08b7f7cc 812 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
813 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
814 if ($iLastMatch !== $j) {
bcae324b 815 if ($htmlsafe) {
9f7f68c3 816 $ret .= '&#32;';
08b7f7cc 817 } else {
818 $ret .= ' ';
819 }
820 }
821 $iLastMatch = $i;
822 $j = $i;
bcae324b 823 if ($htmlsafe) {
cb718de0 824 $ret .= htmlspecialchars($res[1]);
825 } else {
826 $ret .= $res[1];
827 }
098ea084 828 $encoding = ucfirst($res[3]);
d6f584fc 829
830 /* decide about valid decoding */
831 if ($decide && is_conversion_safe($res[2])) {
8bd0068d 832 $utfencode=true;
833 $can_be_encoded=true;
d6f584fc 834 } else {
8bd0068d 835 $can_be_encoded=false;
d6f584fc 836 }
098ea084 837 switch ($encoding)
838 {
8bd0068d 839 case 'B':
840 $replace = base64_decode($res[4]);
841 if ($utfencode) {
842 if ($can_be_encoded) {
843 /* convert string to different charset,
844 * if functions asks for it (usually in compose)
845 */
bcae324b 846 $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsafe);
8bd0068d 847 } else {
848 // convert string to html codes in order to display it
849 $ret .= charset_decode($res[2],$replace);
850 }
fab65ca9 851 } else {
bcae324b 852 if ($htmlsafe) {
8bd0068d 853 $replace = htmlspecialchars($replace);
854 }
855 $ret.= $replace;
fab65ca9 856 }
8bd0068d 857 break;
858 case 'Q':
859 $replace = str_replace('_', ' ', $res[4]);
860 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
861 $replace);
862 if ($utfencode) {
863 if ($can_be_encoded) {
864 /* convert string to different charset,
865 * if functions asks for it (usually in compose)
866 */
bcae324b 867 $replace = charset_convert($res[2], $replace,$default_charset,$htmlsafe);
8bd0068d 868 } else {
869 // convert string to html codes in order to display it
870 $replace = charset_decode($res[2], $replace);
871 }
872 } else {
bcae324b 873 if ($htmlsafe) {
8bd0068d 874 $replace = htmlspecialchars($replace);
875 }
098ea084 876 }
8bd0068d 877 $ret .= $replace;
878 break;
879 default:
880 break;
79e07c7e 881 }
098ea084 882 $chunk = $res[5];
883 $encoded = true;
08b7f7cc 884 }
885 if (!$encoded) {
bcae324b 886 if ($htmlsafe) {
9f7f68c3 887 $ret .= '&#32;';
08b7f7cc 888 } else {
889 $ret .= ' ';
da2415c1 890 }
08b7f7cc 891 }
dc3d13a7 892
bcae324b 893 if (!$encoded && $htmlsafe) {
dc3d13a7 894 $ret .= htmlspecialchars($chunk);
895 } else {
896 $ret .= $chunk;
897 }
098ea084 898 ++$i;
899 }
fd81e884 900 /* remove the first added space */
901 if ($ret) {
bcae324b 902 if ($htmlsafe) {
9f7f68c3 903 $ret = substr($ret,5);
fd81e884 904 } else {
905 $ret = substr($ret,1);
906 }
907 }
da2415c1 908
08b7f7cc 909 return $ret;
451f74a2 910}
911
9f7f68c3 912/**
a24cf710 913 * Encodes header
8bd0068d 914 *
a24cf710 915 * Function uses XTRA_CODE _encodeheader function, if such function exists.
a24cf710 916 *
758a7889 917 * Function uses Q encoding by default and encodes a string according to RFC
918 * 1522 for use in headers if it contains 8-bit characters or anything that
a24cf710 919 * looks like it should be encoded.
8bd0068d 920 *
758a7889 921 * Function switches to B encoding and encodeHeaderBase64() function, if
922 * string is 8bit and multibyte character set supported by mbstring extension
923 * is used. It can cause E_USER_NOTICE errors, if interface is used with
f270a6eb 924 * multibyte character set unsupported by mbstring extension.
925 *
8bd0068d 926 * @param string $string header string, that has to be encoded
927 * @return string quoted-printable encoded string
f270a6eb 928 * @todo make $mb_charsets system wide constant
8bd0068d 929 */
451f74a2 930function encodeHeader ($string) {
6fbd125b 931 global $default_charset, $languages, $squirrelmail_language;
83be314a 932
10dec454 933 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 934 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
33a55f5a 935 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
83be314a 936 }
793cc001 937
a24cf710 938 // Use B encoding for multibyte charsets
f270a6eb 939 $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
940 if (in_array($default_charset,$mb_charsets) &&
941 in_array($default_charset,sq_mb_list_encodings()) &&
942 sq_is8bit($string)) {
943 return encodeHeaderBase64($string,$default_charset);
944 } elseif (in_array($default_charset,$mb_charsets) &&
945 sq_is8bit($string) &&
946 ! in_array($default_charset,sq_mb_list_encodings())) {
947 // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
948 // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
949 }
a24cf710 950
451f74a2 951 // Encode only if the string contains 8-bit characters or =?
3d8371be 952 $j = strlen($string);
098ea084 953 $max_l = 75 - strlen($default_charset) - 7;
954 $aRet = array();
451f74a2 955 $ret = '';
c96c32f4 956 $iEncStart = $enc_init = false;
0d53f0f9 957 $cur_l = $iOffset = 0;
3d8371be 958 for($i = 0; $i < $j; ++$i) {
c96c32f4 959 switch($string{$i})
960 {
8bd0068d 961 case '=':
962 case '<':
963 case '>':
964 case ',':
965 case '?':
966 case '_':
967 if ($iEncStart === false) {
968 $iEncStart = $i;
969 }
970 $cur_l+=3;
971 if ($cur_l > ($max_l-2)) {
972 /* if there is an stringpart that doesn't need encoding, add it */
08b7f7cc 973 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
c96c32f4 974 $aRet[] = "=?$default_charset?Q?$ret?=";
975 $iOffset = $i;
976 $cur_l = 0;
977 $ret = '';
978 $iEncStart = false;
08b7f7cc 979 } else {
8bd0068d 980 $ret .= sprintf("=%02X",ord($string{$i}));
c96c32f4 981 }
8bd0068d 982 break;
983 case '(':
984 case ')':
985 if ($iEncStart !== false) {
08b7f7cc 986 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
8bd0068d 987 $aRet[] = "=?$default_charset?Q?$ret?=";
c96c32f4 988 $iOffset = $i;
8bd0068d 989 $cur_l = 0;
990 $ret = '';
991 $iEncStart = false;
c96c32f4 992 }
8bd0068d 993 break;
994 case ' ':
c96c32f4 995 if ($iEncStart !== false) {
098ea084 996 $cur_l++;
997 if ($cur_l > $max_l) {
08b7f7cc 998 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
c96c32f4 999 $aRet[] = "=?$default_charset?Q?$ret?=";
c96c32f4 1000 $iOffset = $i;
1001 $cur_l = 0;
098ea084 1002 $ret = '';
8bd0068d 1003 $iEncStart = false;
08b7f7cc 1004 } else {
8bd0068d 1005 $ret .= '_';
c96c32f4 1006 }
3d8371be 1007 }
8bd0068d 1008 break;
1009 default:
1010 $k = ord($string{$i});
1011 if ($k > 126) {
1012 if ($iEncStart === false) {
1013 // do not start encoding in the middle of a string, also take the rest of the word.
1014 $sLeadString = substr($string,0,$i);
1015 $aLeadString = explode(' ',$sLeadString);
1016 $sToBeEncoded = array_pop($aLeadString);
1017 $iEncStart = $i - strlen($sToBeEncoded);
1018 $ret .= $sToBeEncoded;
1019 $cur_l += strlen($sToBeEncoded);
1020 }
1021 $cur_l += 3;
1022 /* first we add the encoded string that reached it's max size */
1023 if ($cur_l > ($max_l-2)) {
1024 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1025 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
1026 $cur_l = 3;
1027 $ret = '';
1028 $iOffset = $i;
1029 $iEncStart = $i;
1030 }
1031 $enc_init = true;
1032 $ret .= sprintf("=%02X", $k);
1033 } else {
1034 if ($iEncStart !== false) {
1035 $cur_l++;
1036 if ($cur_l > $max_l) {
1037 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1038 $aRet[] = "=?$default_charset?Q?$ret?=";
1039 $iEncStart = false;
1040 $iOffset = $i;
1041 $cur_l = 0;
1042 $ret = '';
1043 } else {
1044 $ret .= $string{$i};
1045 }
1046 }
1047 }
1048 break;
f7b3ba37 1049 }
451f74a2 1050 }
793cc001 1051
c96c32f4 1052 if ($enc_init) {
1053 if ($iEncStart !== false) {
1054 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1055 $aRet[] = "=?$default_charset?Q?$ret?=";
1056 } else {
1057 $aRet[] = substr($string,$iOffset);
1058 }
1059 $string = implode('',$aRet);
451f74a2 1060 }
3d8371be 1061 return $string;
451f74a2 1062}
b74ba498 1063
f270a6eb 1064/**
1065 * Encodes string according to rfc2047 B encoding header formating rules
1066 *
758a7889 1067 * It is recommended way to encode headers with character sets that store
f270a6eb 1068 * symbols in more than one byte.
1069 *
1070 * Function requires mbstring support. If required mbstring functions are missing,
1071 * function returns false and sets E_USER_WARNING level error message.
1072 *
758a7889 1073 * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
1074 * that mbstring functions will generate E_WARNING errors, if unsupported
f270a6eb 1075 * character set is used. mb_encode_mimeheader function provided by php
1076 * mbstring extension is not used in order to get better control of header
1077 * encoding.
1078 *
758a7889 1079 * Used php code functions - function_exists(), trigger_error(), strlen()
1080 * (is used with charset names and base64 strings). Used php mbstring
f270a6eb 1081 * functions - mb_strlen and mb_substr.
1082 *
758a7889 1083 * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
f270a6eb 1084 * encoding), rfc 2822 (header folding)
1085 *
1086 * @param string $string header string that must be encoded
758a7889 1087 * @param string $charset character set. Must be supported by mbstring extension.
f270a6eb 1088 * Use sq_mb_list_encodings() to detect supported charsets.
1089 * @return string string encoded according to rfc2047 B encoding formating rules
1090 * @since 1.5.1
1091 * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1092 * @todo Do we want to control max length of header?
1093 * @todo Do we want to control EOL (end-of-line) marker?
1094 * @todo Do we want to translate error message?
1095 */
1096function encodeHeaderBase64($string,$charset) {
1097 /**
1098 * Check mbstring function requirements.
1099 */
1100 if (! function_exists('mb_strlen') ||
1101 ! function_exists('mb_substr')) {
1102 // set E_USER_WARNING
1103 trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
1104 // return false
1105 return false;
1106 }
1107
1108 // initial return array
1109 $aRet = array();
1110
1111 /**
1112 * header length = 75 symbols max (same as in encodeHeader)
1113 * remove $charset length
1114 * remove =? ? ?= (5 chars)
1115 * remove 2 more chars (\r\n ?)
1116 */
1117 $iMaxLength = 75 - strlen($charset) - 7;
1118
1119 // set first character position
1120 $iStartCharNum = 0;
1121
1122 // loop through all characters. count characters and not bytes.
1123 for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1124 // encode string from starting character to current character.
1125 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1126
1127 // Check encoded string length
1128 if(strlen($encoded_string)>$iMaxLength) {
1129 // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1130 $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1131
1132 // set new starting character
1133 $iStartCharNum = $iCharNum-1;
1134
1135 // encode last char (in case it is last character in string)
1136 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1137 } // if string is shorter than max length - add next character
1138 }
1139
1140 // add last encoded string to array
1141 $aRet[] = $encoded_string;
1142
1143 // set initial return string
1144 $sRet = '';
1145
1146 // loop through encoded strings
1147 foreach($aRet as $string) {
1148 // TODO: Do we want to control EOL (end-of-line) marker
1149 if ($sRet!='') $sRet.= " ";
1150
1151 // add header tags and encoded string to return string
1152 $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1153 }
1154
1155 return $sRet;
1156}
1157
691a2d25 1158/* This function trys to locate the entity_id of a specific mime element */
3d8371be 1159function find_ent_id($id, $message) {
a171b359 1160 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1161 if ($message->entities[$i]->header->type0 == 'multipart') {
3d8371be 1162 $ret = find_ent_id($id, $message->entities[$i]);
451f74a2 1163 } else {
3d8371be 1164 if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
d8cffbab 1165// if (sq_check_save_extension($message->entities[$i])) {
8bd0068d 1166 return $message->entities[$i]->entity_id;
da2415c1 1167// }
c8f5f606 1168 } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1169 /**
1170 * This is part of a fix for Outlook Express 6.x generating
1171 * cid URLs without creating content-id headers
1172 * @@JA - 20050207
1173 */
1174 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1175 return $message->entities[$i]->entity_id;
1176 }
3d8371be 1177 }
a3daaaf3 1178 }
451f74a2 1179 }
3d8371be 1180 return $ret;
451f74a2 1181}
a3daaaf3 1182
e5e9381a 1183function sq_check_save_extension($message) {
1184 $filename = $message->getFilename();
1185 $ext = substr($filename, strrpos($filename,'.')+1);
1186 $save_extensions = array('jpg','jpeg','gif','png','bmp');
3d8371be 1187 return in_array($ext, $save_extensions);
e5e9381a 1188}
1189
1190
691a2d25 1191/**
8bd0068d 1192 ** HTMLFILTER ROUTINES
1193 */
451f74a2 1194
691a2d25 1195/**
0493ed11 1196 * This function checks attribute values for entity-encoded values
1197 * and returns them translated into 8-bit strings so we can run
1198 * checks on them.
8bd0068d 1199 *
0493ed11 1200 * @param $attvalue A string to run entity check against.
1201 * @return Nothing, modifies a reference value.
8bd0068d 1202 */
0493ed11 1203function sq_defang(&$attvalue){
1204 $me = 'sq_defang';
2dd879b8 1205 /**
0493ed11 1206 * Skip this if there aren't ampersands or backslashes.
8bd0068d 1207 */
0493ed11 1208 if (strpos($attvalue, '&') === false
1209 && strpos($attvalue, '\\') === false){
1210 return;
2dd879b8 1211 }
0493ed11 1212 $m = false;
2b665f28 1213 // before deent, translate the dangerous unicode characters and ... to safe values
1214 // otherwise the regular expressions do not match.
1215
1216
1217
0493ed11 1218 do {
1219 $m = false;
1220 $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1221 $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1222 $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1223 } while ($m == true);
1224 $attvalue = stripslashes($attvalue);
2dd879b8 1225}
1226
1227/**
8bd0068d 1228 * Kill any tabs, newlines, or carriage returns. Our friends the
1229 * makers of the browser with 95% market value decided that it'd
1230 * be funny to make "java[tab]script" be just as good as "javascript".
1231 *
1232 * @param attvalue The attribute value before extraneous spaces removed.
0493ed11 1233 * @return attvalue Nothing, modifies a reference value.
8bd0068d 1234 */
0493ed11 1235function sq_unspace(&$attvalue){
1236 $me = 'sq_unspace';
1237 if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1238 $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1239 Array('', '', '', '', ''), $attvalue);
2dd879b8 1240 }
2dd879b8 1241}
1242
1243/**
88de4926 1244 * Translate all dangerous Unicode or Shift_JIS characters which are accepted by
2b665f28 1245 * IE as regular characters.
1246 *
1247 * @param attvalue The attribute value before dangerous characters are translated.
1248 * @return attvalue Nothing, modifies a reference value.
1249 * @author Marc Groot Koerkamp.
1250 */
1251function sq_fixIE_idiocy(&$attvalue) {
1252 // remove NUL
1253 $attvalue = str_replace("\0", "", $attvalue);
1254 // remove comments
1255 $attvalue = preg_replace("/(\/\*.*?\*\/)/","",$attvalue);
1256
88de4926 1257 // IE has the evil habit of accepting every possible value for the attribute expression.
1258 // The table below contains characters which are parsed by IE if they are used in the "expression"
2b665f28 1259 // attribute value.
1260 $aDangerousCharsReplacementTable = array(
1261 array('&#x029F;', '&#0671;' ,/* L UNICODE IPA Extension */
1262 '&#x0280;', '&#0640;' ,/* R UNICODE IPA Extension */
1263 '&#x0274;', '&#0628;' ,/* N UNICODE IPA Extension */
567dc524 1264 '&#xFF25;', '&#65317;' ,/* Unicode FULLWIDTH LATIN CAPITAL LETTER E */
1265 '&#xFF45;', '&#65349;' ,/* Unicode FULLWIDTH LATIN SMALL LETTER E */
2b665f28 1266 '&#xFF38;', '&#65336;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER X */
1267 '&#xFF58;', '&#65368;',/* Unicode FULLWIDTH LATIN SMALL LETTER X */
1268 '&#xFF30;', '&#65328;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER P */
1269 '&#xFF50;', '&#65360;',/* Unicode FULLWIDTH LATIN SMALL LETTER P */
1270 '&#xFF32;', '&#65330;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER R */
1271 '&#xFF52;', '&#65362;',/* Unicode FULLWIDTH LATIN SMALL LETTER R */
1272 '&#xFF33;', '&#65331;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER S */
1273 '&#xFF53;', '&#65363;',/* Unicode FULLWIDTH LATIN SMALL LETTER S */
1274 '&#xFF29;', '&#65321;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER I */
1275 '&#xFF49;', '&#65353;',/* Unicode FULLWIDTH LATIN SMALL LETTER I */
1276 '&#xFF2F;', '&#65327;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER O */
1277 '&#xFF4F;', '&#65359;',/* Unicode FULLWIDTH LATIN SMALL LETTER O */
1278 '&#xFF2E;', '&#65326;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER N */
1279 '&#xFF4E;', '&#65358;',/* Unicode FULLWIDTH LATIN SMALL LETTER N */
1280 '&#xFF2C;', '&#65324;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER L */
1281 '&#xFF4C;', '&#65356;',/* Unicode FULLWIDTH LATIN SMALL LETTER L */
1282 '&#xFF35;', '&#65333;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER U */
1283 '&#xFF55;', '&#65365;',/* Unicode FULLWIDTH LATIN SMALL LETTER U */
1284 '&#x207F;', '&#8319;' ,/* Unicode SUPERSCRIPT LATIN SMALL LETTER N */
567dc524 1285 "\xEF\xBC\xA5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER E */ // in unicode this is some Chinese char range
1286 "\xEF\xBD\x85", /* Shift JIS FULLWIDTH LATIN SMALL LETTER E */
1287 "\xEF\xBC\xB8", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER X */
1288 "\xEF\xBD\x98", /* Shift JIS FULLWIDTH LATIN SMALL LETTER X */
1289 "\xEF\xBC\xB0", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER P */
1290 "\xEF\xBD\x90", /* Shift JIS FULLWIDTH LATIN SMALL LETTER P */
1291 "\xEF\xBC\xB2", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER R */
1292 "\xEF\xBD\x92", /* Shift JIS FULLWIDTH LATIN SMALL LETTER R */
1293 "\xEF\xBC\xB3", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER S */
1294 "\xEF\xBD\x93", /* Shift JIS FULLWIDTH LATIN SMALL LETTER S */
1295 "\xEF\xBC\xA9", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER I */
1296 "\xEF\xBD\x89", /* Shift JIS FULLWIDTH LATIN SMALL LETTER I */
1297 "\xEF\xBC\xAF", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER O */
1298 "\xEF\xBD\x8F", /* Shift JIS FULLWIDTH LATIN SMALL LETTER O */
1299 "\xEF\xBC\xAE", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER N */
1300 "\xEF\xBD\x8E", /* Shift JIS FULLWIDTH LATIN SMALL LETTER N */
1301 "\xEF\xBC\xAC", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER L */
1302 "\xEF\xBD\x8C", /* Shift JIS FULLWIDTH LATIN SMALL LETTER L */
1303 "\xEF\xBC\xB5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER U */
1304 "\xEF\xBD\x95", /* Shift JIS FULLWIDTH LATIN SMALL LETTER U */
1305 "\xE2\x81\xBF", /* Shift JIS FULLWIDTH SUPERSCRIPT N */
1306 "\xCA\x9F", /* L UNICODE IPA Extension */
1307 "\xCA\x80", /* R UNICODE IPA Extension */
1308 "\xC9\xB4"), /* N UNICODE IPA Extension */
2b665f28 1309 array('l', 'l', 'r','r','n','n',
567dc524 1310 'E','E','e','e','X','X','x','x','P','P','p','p','R','R','r','r','S','S','s','s','I','I',
1311 'i','i','O','O','o','o','N','N','n','n','L','L','l','l','U','U','u','u','n','n',
1312 '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 1313 $attvalue = str_replace($aDangerousCharsReplacementTable[0],$aDangerousCharsReplacementTable[1],$attvalue);
1314
88de4926 1315 // Escapes are useful for special characters like "{}[]()'&. In other cases they are
1316 // used for XSS.
2b665f28 1317 $attvalue = preg_replace("/(\\\\)([a-zA-Z]{1})/",'$2',$attvalue);
1318}
1319
1320/**
8bd0068d 1321 * This function returns the final tag out of the tag name, an array
1322 * of attributes, and the type of the tag. This function is called by
1323 * sq_sanitize internally.
1324 *
1325 * @param $tagname the name of the tag.
1326 * @param $attary the array of attributes and their values
1327 * @param $tagtype The type of the tag (see in comments).
1328 * @return a string with the final tag representation.
1329 */
691a2d25 1330function sq_tagprint($tagname, $attary, $tagtype){
b583c3e8 1331 $me = 'sq_tagprint';
3d8371be 1332
691a2d25 1333 if ($tagtype == 2){
1334 $fulltag = '</' . $tagname . '>';
1335 } else {
1336 $fulltag = '<' . $tagname;
1337 if (is_array($attary) && sizeof($attary)){
1338 $atts = Array();
1339 while (list($attname, $attvalue) = each($attary)){
1340 array_push($atts, "$attname=$attvalue");
1341 }
1342 $fulltag .= ' ' . join(" ", $atts);
1343 }
1344 if ($tagtype == 3){
b583c3e8 1345 $fulltag .= ' /';
691a2d25 1346 }
b583c3e8 1347 $fulltag .= '>';
451f74a2 1348 }
691a2d25 1349 return $fulltag;
451f74a2 1350}
a3daaaf3 1351
691a2d25 1352/**
8bd0068d 1353 * A small helper function to use with array_walk. Modifies a by-ref
1354 * value and makes it lowercase.
1355 *
1356 * @param $val a value passed by-ref.
1357 * @return void since it modifies a by-ref value.
1358 */
691a2d25 1359function sq_casenormalize(&$val){
1360 $val = strtolower($val);
1361}
451f74a2 1362
691a2d25 1363/**
8bd0068d 1364 * This function skips any whitespace from the current position within
1365 * a string and to the next non-whitespace value.
1366 *
1367 * @param $body the string
1368 * @param $offset the offset within the string where we should start
1369 * looking for the next non-whitespace character.
1370 * @return the location within the $body where the next
1371 * non-whitespace char is located.
1372 */
691a2d25 1373function sq_skipspace($body, $offset){
b583c3e8 1374 $me = 'sq_skipspace';
3d8371be 1375 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
691a2d25 1376 if (sizeof($matches{1})){
1377 $count = strlen($matches{1});
1378 $offset += $count;
451f74a2 1379 }
691a2d25 1380 return $offset;
451f74a2 1381}
a3daaaf3 1382
691a2d25 1383/**
8bd0068d 1384 * This function looks for the next character within a string. It's
1385 * really just a glorified "strpos", except it catches if failures
1386 * nicely.
1387 *
1388 * @param $body The string to look for needle in.
1389 * @param $offset Start looking from this position.
1390 * @param $needle The character/string to look for.
1391 * @return location of the next occurance of the needle, or
1392 * strlen($body) if needle wasn't found.
1393 */
691a2d25 1394function sq_findnxstr($body, $offset, $needle){
3d8371be 1395 $me = 'sq_findnxstr';
691a2d25 1396 $pos = strpos($body, $needle, $offset);
1397 if ($pos === FALSE){
1398 $pos = strlen($body);
451f74a2 1399 }
691a2d25 1400 return $pos;
451f74a2 1401}
a3daaaf3 1402
691a2d25 1403/**
8bd0068d 1404 * This function takes a PCRE-style regexp and tries to match it
1405 * within the string.
1406 *
1407 * @param $body The string to look for needle in.
1408 * @param $offset Start looking from here.
1409 * @param $reg A PCRE-style regex to match.
1410 * @return Returns a false if no matches found, or an array
1411 * with the following members:
1412 * - integer with the location of the match within $body
1413 * - string with whatever content between offset and the match
1414 * - string with whatever it is we matched
1415 */
691a2d25 1416function sq_findnxreg($body, $offset, $reg){
b583c3e8 1417 $me = 'sq_findnxreg';
691a2d25 1418 $matches = Array();
1419 $retarr = Array();
7d06541f 1420 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1421 if (!isset($matches{0}) || !$matches{0}){
691a2d25 1422 $retarr = false;
1423 } else {
1424 $retarr{0} = $offset + strlen($matches{1});
1425 $retarr{1} = $matches{1};
1426 $retarr{2} = $matches{2};
1427 }
1428 return $retarr;
1429}
a3daaaf3 1430
691a2d25 1431/**
8bd0068d 1432 * This function looks for the next tag.
1433 *
1434 * @param $body String where to look for the next tag.
1435 * @param $offset Start looking from here.
1436 * @return false if no more tags exist in the body, or
1437 * an array with the following members:
1438 * - string with the name of the tag
1439 * - array with attributes and their values
1440 * - integer with tag type (1, 2, or 3)
1441 * - integer where the tag starts (starting "<")
1442 * - integer where the tag ends (ending ">")
1443 * first three members will be false, if the tag is invalid.
1444 */
691a2d25 1445function sq_getnxtag($body, $offset){
b583c3e8 1446 $me = 'sq_getnxtag';
691a2d25 1447 if ($offset > strlen($body)){
1448 return false;
1449 }
1450 $lt = sq_findnxstr($body, $offset, "<");
1451 if ($lt == strlen($body)){
1452 return false;
1453 }
1454 /**
8bd0068d 1455 * We are here:
1456 * blah blah <tag attribute="value">
1457 * \---------^
1458 */
691a2d25 1459 $pos = sq_skipspace($body, $lt+1);
1460 if ($pos >= strlen($body)){
1461 return Array(false, false, false, $lt, strlen($body));
1462 }
1463 /**
8bd0068d 1464 * There are 3 kinds of tags:
1465 * 1. Opening tag, e.g.:
1466 * <a href="blah">
1467 * 2. Closing tag, e.g.:
1468 * </a>
1469 * 3. XHTML-style content-less tag, e.g.:
1470 * <img src="blah" />
1471 */
691a2d25 1472 $tagtype = false;
1473 switch (substr($body, $pos, 1)){
3d8371be 1474 case '/':
1475 $tagtype = 2;
1476 $pos++;
1477 break;
1478 case '!':
1479 /**
8bd0068d 1480 * A comment or an SGML declaration.
1481 */
3d8371be 1482 if (substr($body, $pos+1, 2) == "--"){
1483 $gt = strpos($body, "-->", $pos);
1484 if ($gt === false){
1485 $gt = strlen($body);
1486 } else {
1487 $gt += 2;
1488 }
1489 return Array(false, false, false, $lt, $gt);
bb8d0799 1490 } else {
3d8371be 1491 $gt = sq_findnxstr($body, $pos, ">");
1492 return Array(false, false, false, $lt, $gt);
1493 }
1494 break;
1495 default:
1496 /**
8bd0068d 1497 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1498 * later.
1499 */
3d8371be 1500 $tagtype = 1;
1501 break;
691a2d25 1502 }
a3daaaf3 1503
0493ed11 1504 $tag_start = $pos;
691a2d25 1505 $tagname = '';
1506 /**
8bd0068d 1507 * Look for next [\W-_], which will indicate the end of the tag name.
1508 */
691a2d25 1509 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1510 if ($regary == false){
1511 return Array(false, false, false, $lt, strlen($body));
1512 }
1513 list($pos, $tagname, $match) = $regary;
1514 $tagname = strtolower($tagname);
1515
1516 /**
8bd0068d 1517 * $match can be either of these:
1518 * '>' indicating the end of the tag entirely.
1519 * '\s' indicating the end of the tag name.
1520 * '/' indicating that this is type-3 xhtml tag.
1521 *
1522 * Whatever else we find there indicates an invalid tag.
1523 */
691a2d25 1524 switch ($match){
3d8371be 1525 case '/':
691a2d25 1526 /**
8bd0068d 1527 * This is an xhtml-style tag with a closing / at the
1528 * end, like so: <img src="blah" />. Check if it's followed
1529 * by the closing bracket. If not, then this tag is invalid
1530 */
3d8371be 1531 if (substr($body, $pos, 2) == "/>"){
1532 $pos++;
1533 $tagtype = 3;
1534 } else {
1535 $gt = sq_findnxstr($body, $pos, ">");
1536 $retary = Array(false, false, false, $lt, $gt);
1537 return $retary;
1538 }
1539 case '>':
1540 return Array($tagname, false, $tagtype, $lt, $pos);
1541 break;
1542 default:
1543 /**
8bd0068d 1544 * Check if it's whitespace
1545 */
3d8371be 1546 if (!preg_match('/\s/', $match)){
1547 /**
8bd0068d 1548 * This is an invalid tag! Look for the next closing ">".
1549 */
7d06541f 1550 $gt = sq_findnxstr($body, $lt, ">");
3d8371be 1551 return Array(false, false, false, $lt, $gt);
1552 }
1553 break;
691a2d25 1554 }
3d8371be 1555
691a2d25 1556 /**
8bd0068d 1557 * At this point we're here:
1558 * <tagname attribute='blah'>
1559 * \-------^
1560 *
1561 * At this point we loop in order to find all attributes.
1562 */
691a2d25 1563 $attname = '';
0493ed11 1564 $atttype = false;
691a2d25 1565 $attary = Array();
1566
1567 while ($pos <= strlen($body)){
1568 $pos = sq_skipspace($body, $pos);
1569 if ($pos == strlen($body)){
1570 /**
8bd0068d 1571 * Non-closed tag.
1572 */
691a2d25 1573 return Array(false, false, false, $lt, $pos);
1574 }
1575 /**
8bd0068d 1576 * See if we arrived at a ">" or "/>", which means that we reached
1577 * the end of the tag.
1578 */
691a2d25 1579 $matches = Array();
164800ad 1580 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
c828931c 1581 /**
8bd0068d 1582 * Yep. So we did.
1583 */
c828931c 1584 $pos += strlen($matches{1});
1585 if ($matches{2} == "/>"){
1586 $tagtype = 3;
1587 $pos++;
1588 }
1589 return Array($tagname, $attary, $tagtype, $lt, $pos);
1590 }
a3daaaf3 1591
cca46357 1592 /**
8bd0068d 1593 * There are several types of attributes, with optional
1594 * [:space:] between members.
1595 * Type 1:
1596 * attrname[:space:]=[:space:]'CDATA'
1597 * Type 2:
1598 * attrname[:space:]=[:space:]"CDATA"
1599 * Type 3:
1600 * attr[:space:]=[:space:]CDATA
1601 * Type 4:
1602 * attrname
1603 *
1604 * We leave types 1 and 2 the same, type 3 we check for
1605 * '"' and convert to "&quot" if needed, then wrap in
1606 * double quotes. Type 4 we convert into:
1607 * attrname="yes".
1608 */
3f7c623f 1609 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
691a2d25 1610 if ($regary == false){
1611 /**
8bd0068d 1612 * Looks like body ended before the end of tag.
1613 */
691a2d25 1614 return Array(false, false, false, $lt, strlen($body));
cca46357 1615 }
691a2d25 1616 list($pos, $attname, $match) = $regary;
1617 $attname = strtolower($attname);
1618 /**
8bd0068d 1619 * We arrived at the end of attribute name. Several things possible
1620 * here:
1621 * '>' means the end of the tag and this is attribute type 4
1622 * '/' if followed by '>' means the same thing as above
1623 * '\s' means a lot of things -- look what it's followed by.
1624 * anything else means the attribute is invalid.
1625 */
691a2d25 1626 switch($match){
3d8371be 1627 case '/':
691a2d25 1628 /**
8bd0068d 1629 * This is an xhtml-style tag with a closing / at the
1630 * end, like so: <img src="blah" />. Check if it's followed
1631 * by the closing bracket. If not, then this tag is invalid
1632 */
3d8371be 1633 if (substr($body, $pos, 2) == "/>"){
691a2d25 1634 $pos++;
3d8371be 1635 $tagtype = 3;
691a2d25 1636 } else {
3d8371be 1637 $gt = sq_findnxstr($body, $pos, ">");
1638 $retary = Array(false, false, false, $lt, $gt);
1639 return $retary;
1640 }
1641 case '>':
1642 $attary{$attname} = '"yes"';
1643 return Array($tagname, $attary, $tagtype, $lt, $pos);
1644 break;
1645 default:
1646 /**
8bd0068d 1647 * Skip whitespace and see what we arrive at.
1648 */
3d8371be 1649 $pos = sq_skipspace($body, $pos);
1650 $char = substr($body, $pos, 1);
1651 /**
8bd0068d 1652 * Two things are valid here:
1653 * '=' means this is attribute type 1 2 or 3.
1654 * \w means this was attribute type 4.
1655 * anything else we ignore and re-loop. End of tag and
1656 * invalid stuff will be caught by our checks at the beginning
1657 * of the loop.
1658 */
3d8371be 1659 if ($char == "="){
1660 $pos++;
1661 $pos = sq_skipspace($body, $pos);
691a2d25 1662 /**
8bd0068d 1663 * Here are 3 possibilities:
1664 * "'" attribute type 1
1665 * '"' attribute type 2
1666 * everything else is the content of tag type 3
1667 */
3d8371be 1668 $quot = substr($body, $pos, 1);
1669 if ($quot == "'"){
1670 $regary = sq_findnxreg($body, $pos+1, "\'");
1671 if ($regary == false){
1672 return Array(false, false, false, $lt, strlen($body));
1673 }
1674 list($pos, $attval, $match) = $regary;
1675 $pos++;
1676 $attary{$attname} = "'" . $attval . "'";
1677 } else if ($quot == '"'){
1678 $regary = sq_findnxreg($body, $pos+1, '\"');
1679 if ($regary == false){
1680 return Array(false, false, false, $lt, strlen($body));
1681 }
1682 list($pos, $attval, $match) = $regary;
1683 $pos++;
1684 $attary{$attname} = '"' . $attval . '"';
1685 } else {
1686 /**
8bd0068d 1687 * These are hateful. Look for \s, or >.
1688 */
3d8371be 1689 $regary = sq_findnxreg($body, $pos, "[\s>]");
1690 if ($regary == false){
1691 return Array(false, false, false, $lt, strlen($body));
1692 }
1693 list($pos, $attval, $match) = $regary;
1694 /**
8bd0068d 1695 * If it's ">" it will be caught at the top.
1696 */
3d8371be 1697 $attval = preg_replace("/\"/s", "&quot;", $attval);
1698 $attary{$attname} = '"' . $attval . '"';
7e235a1a 1699 }
3d8371be 1700 } else if (preg_match("|[\w/>]|", $char)) {
691a2d25 1701 /**
8bd0068d 1702 * That was attribute type 4.
1703 */
3d8371be 1704 $attary{$attname} = '"yes"';
1705 } else {
1706 /**
8bd0068d 1707 * An illegal character. Find next '>' and return.
1708 */
3d8371be 1709 $gt = sq_findnxstr($body, $pos, ">");
1710 return Array(false, false, false, $lt, $gt);
451f74a2 1711 }
3d8371be 1712 break;
691a2d25 1713 }
1714 }
1715 /**
8bd0068d 1716 * The fact that we got here indicates that the tag end was never
1717 * found. Return invalid tag indication so it gets stripped.
1718 */
691a2d25 1719 return Array(false, false, false, $lt, strlen($body));
1720}
1721
1722/**
0493ed11 1723 * Translates entities into literal values so they can be checked.
8bd0068d 1724 *
0493ed11 1725 * @param $attvalue the by-ref value to check.
1726 * @param $regex the regular expression to check against.
1727 * @param $hex whether the entites are hexadecimal.
1728 * @return True or False depending on whether there were matches.
8bd0068d 1729 */
0493ed11 1730function sq_deent(&$attvalue, $regex, $hex=false){
b583c3e8 1731 $me = 'sq_deent';
0493ed11 1732 $ret_match = false;
2b665f28 1733 // remove comments
1734 //$attvalue = preg_replace("/(\/\*.*\*\/)/","",$attvalue);
0493ed11 1735 preg_match_all($regex, $attvalue, $matches);
1736 if (is_array($matches) && sizeof($matches[0]) > 0){
1737 $repl = Array();
1738 for ($i = 0; $i < sizeof($matches[0]); $i++){
1739 $numval = $matches[1][$i];
1740 if ($hex){
1741 $numval = hexdec($numval);
a3daaaf3 1742 }
0493ed11 1743 $repl{$matches[0][$i]} = chr($numval);
691a2d25 1744 }
0493ed11 1745 $attvalue = strtr($attvalue, $repl);
1746 return true;
1747 } else {
1748 return false;
691a2d25 1749 }
691a2d25 1750}
1751
1752/**
8bd0068d 1753 * This function runs various checks against the attributes.
1754 *
1755 * @param $tagname String with the name of the tag.
1756 * @param $attary Array with all tag attributes.
1757 * @param $rm_attnames See description for sq_sanitize
1758 * @param $bad_attvals See description for sq_sanitize
1759 * @param $add_attr_to_tag See description for sq_sanitize
1760 * @param $message message object
1761 * @param $id message id
1762 * @return Array with modified attributes.
1763 */
da2415c1 1764function sq_fixatts($tagname,
1765 $attary,
691a2d25 1766 $rm_attnames,
1767 $bad_attvals,
1768 $add_attr_to_tag,
1769 $message,
b3af12ef 1770 $id,
3d8371be 1771 $mailbox
691a2d25 1772 ){
b583c3e8 1773 $me = 'sq_fixatts';
691a2d25 1774 while (list($attname, $attvalue) = each($attary)){
1775 /**
8bd0068d 1776 * See if this attribute should be removed.
1777 */
691a2d25 1778 foreach ($rm_attnames as $matchtag=>$matchattrs){
1779 if (preg_match($matchtag, $tagname)){
1780 foreach ($matchattrs as $matchattr){
1781 if (preg_match($matchattr, $attname)){
1782 unset($attary{$attname});
1783 continue;
1784 }
451f74a2 1785 }
451f74a2 1786 }
691a2d25 1787 }
1788 /**
2b665f28 1789 * Workaround for IE quirks
1790 */
1791 sq_fixIE_idiocy($attvalue);
1792
1793 /**
8bd0068d 1794 * Remove any backslashes, entities, and extraneous whitespace.
1795 */
2b665f28 1796
1797 $oldattvalue = $attvalue;
0493ed11 1798 sq_defang($attvalue);
2b665f28 1799 if ($attname == 'style' && $attvalue !== $oldattvalue) {
1800 // entities are used in the attribute value. In 99% of the cases it's there as XSS
1801 // i.e.<div style="{ left:exp&#x0280;essio&#x0274;( alert('XSS') ) }">
1802 $attvalue = "idiocy";
1803 $attary{$attname} = $attvalue;
1804 }
0493ed11 1805 sq_unspace($attvalue);
af861a34 1806
1807 /**
8bd0068d 1808 * Now let's run checks on the attvalues.
1809 * I don't expect anyone to comprehend this. If you do,
1810 * get in touch with me so I can drive to where you live and
1811 * shake your hand personally. :)
1812 */
691a2d25 1813 foreach ($bad_attvals as $matchtag=>$matchattrs){
1814 if (preg_match($matchtag, $tagname)){
1815 foreach ($matchattrs as $matchattr=>$valary){
1816 if (preg_match($matchattr, $attname)){
1817 /**
8bd0068d 1818 * There are two arrays in valary.
1819 * First is matches.
1820 * Second one is replacements
1821 */
691a2d25 1822 list($valmatch, $valrepl) = $valary;
da2415c1 1823 $newvalue =
691a2d25 1824 preg_replace($valmatch, $valrepl, $attvalue);
1825 if ($newvalue != $attvalue){
1826 $attary{$attname} = $newvalue;
567dc524 1827 $attvalue = $newvalue;
691a2d25 1828 }
1829 }
1830 }
451f74a2 1831 }
a3daaaf3 1832 }
567dc524 1833 if ($attname == 'style') {
1834 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1835 // 8bit and control characters in style attribute values can be used for XSS, remove them
1836 $attary{$attname} = '"disallowed character"';
1837 }
1838 preg_match_all("/url\s*\((.+)\)/si",$attvalue,$aMatch);
1839 if (count($aMatch)) {
1840 foreach($aMatch[1] as $sMatch) {
1841 // url value
1842 $urlvalue = $sMatch;
1843 sq_fix_url($attname, $urlvalue, $message, $id, $mailbox,"'");
1844 $attary{$attname} = str_replace($sMatch,$urlvalue,$attvalue);
1845 }
1846 }
691a2d25 1847 }
ff940ebc 1848 /**
567dc524 1849 * Use white list based filtering on attributes which can contain url's
ff940ebc 1850 */
567dc524 1851 else if ($attname == 'href' || $attname == 'src' || $attname == 'background') {
1852 sq_fix_url($attname, $attvalue, $message, $id, $mailbox);
1853 $attary{$attname} = $attvalue;
ff940ebc 1854 }
a3daaaf3 1855 }
691a2d25 1856 /**
8bd0068d 1857 * See if we need to append any attributes to this tag.
1858 */
691a2d25 1859 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1860 if (preg_match($matchtag, $tagname)){
1861 $attary = array_merge($attary, $addattary);
1862 }
1863 }
1864 return $attary;
451f74a2 1865}
a3daaaf3 1866
691a2d25 1867/**
567dc524 1868 * This function filters url's
1869 *
1870 * @param $attvalue String with attribute value to filter
1871 * @param $message message object
1872 * @param $id message id
1873 * @param $mailbox mailbox
1874 * @param $sQuote quoting characters around url's
1875 */
1876function sq_fix_url($attname, &$attvalue, $message, $id, $mailbox,$sQuote = '"') {
1877 $attvalue = trim($attvalue);
1878 if ($attvalue && ($attvalue[0] =='"'|| $attvalue[0] == "'")) {
1879 // remove the double quotes
1880 $sQuote = $attvalue[0];
1881 $attvalue = trim(substr($attvalue,1,-1));
1882 }
1883
d31f73f1 1884 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
1885 // images off by default.
afbf184c 1886 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET, FALSE);
d31f73f1 1887
567dc524 1888 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1889
1890 /**
1891 * Replace empty src tags with the blank image. src is only used
1892 * for frames, images, and image inputs. Doing a replace should
1893 * not affect them working as should be, however it will stop
1894 * IE from being kicked off when src for img tags are not set
1895 */
1896 if ($attvalue == '') {
1897 $attvalue = '"' . SM_PATH . 'images/blank.png"';
1898 } else {
1899 // first, disallow 8 bit characters and control characters
1900 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1901 switch ($attname) {
1902 case 'href':
1903 $attvalue = $sQuote . 'http://invalid-stuff-detected.example.com' . $sQuote;
1904 break;
1905 default:
1906 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1907 break;
1908 }
1909 } else {
1910 $aUrl = parse_url($attvalue);
1911 if (isset($aUrl['scheme'])) {
1912 switch(strtolower($aUrl['scheme'])) {
d75e755b 1913 case 'mailto':
567dc524 1914 case 'http':
1915 case 'https':
1916 case 'ftp':
1917 if ($attname != 'href') {
1918 if ($view_unsafe_images == false) {
1919 $attvalue = $sQuote . $secremoveimg . $sQuote;
1920 } else {
1921 if (isset($aUrl['path'])) {
8bbbc757 1922
1923 // No one has been able to show that image URIs
1924 // can be exploited, so for now, no restrictions
1925 // are made at all. If this proves to be a problem,
1926 // the commented-out code below can be of help.
1927 // (One consideration is that I see nothing in this
1928 // function that specifically says that we will
1929 // only ever arrive here when inspecting an image
1930 // tag, although that does seem to be the end
1931 // result - e.g., <script src="..."> where malicious
1932 // image URIs are in fact a problem are already
1933 // filtered out elsewhere.
1934 /* ---------------------------------
567dc524 1935 // validate image extension.
1936 $ext = strtolower(substr($aUrl['path'],strrpos($aUrl['path'],'.')));
1937 if (!in_array($ext,array('.jpeg','.jpg','xjpeg','.gif','.bmp','.jpe','.png','.xbm'))) {
8bbbc757 1938 // If URI is to something other than
1939 // a regular image file, get the contents
1940 // and try to see if it is an image.
1941 // Don't use Fileinfo (finfo_file()) because
1942 // we'd need to make the admin configure the
1943 // location of the magic.mime file (FIXME: add finfo_file() support later?)
1944 //
1945 $mime_type = '';
1946 if (function_exists('mime_content_type')
1947 && ($FILE = @fopen($attvalue, 'rb', FALSE))) {
1948
1949 // fetch file
1950 //
1951 $file_contents = '';
1952 while (!feof($FILE)) {
1953 $file_contents .= fread($FILE, 8192);
1954 }
1955 fclose($FILE);
1956
1957 // store file locally
1958 //
1959 global $attachment_dir, $username;
1960 $hashed_attachment_dir = getHashedDir($username, $attachment_dir);
1961 $localfilename = GenerateRandomString(32, '', 7);
1962 $full_localfilename = "$hashed_attachment_dir/$localfilename";
1963 while (file_exists($full_localfilename)) {
1964 $localfilename = GenerateRandomString(32, '', 7);
1965 $full_localfilename = "$hashed_attachment_dir/$localfilename";
1966 }
1967 $FILE = fopen("$hashed_attachment_dir/$localfilename", 'wb');
1968 fwrite($FILE, $file_contents);
1969 fclose($FILE);
1970
1971 // get mime type and remove file
1972 //
1973 $mime_type = mime_content_type("$hashed_attachment_dir/$localfilename");
1974 unlink("$hashed_attachment_dir/$localfilename");
1975 }
1976 // debug: echo "$attvalue FILE TYPE IS $mime_type<HR>";
1977 if (substr(strtolower($mime_type), 0, 5) != 'image') {
1978 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1979 }
567dc524 1980 }
8bbbc757 1981 --------------------------------- */
567dc524 1982 } else {
1983 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1984 }
1985 }
cf088f55 1986 } else {
1987 $attvalue = $sQuote . $attvalue . $sQuote;
567dc524 1988 }
1989 break;
1990 case 'outbind':
1991 /**
1992 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1993 * One day MS might actually make it match something useful, for now, falling
1994 * back to using cid2http, so we can grab the blank.png.
1995 */
cf088f55 1996 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
567dc524 1997 break;
1998 case 'cid':
1999 /**
2000 * Turn cid: urls into http-friendly ones.
2001 */
cf088f55 2002 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
567dc524 2003 break;
2004 default:
2005 $attvalue = $sQuote . SM_PATH . 'images/blank.png' . $sQuote;
2006 break;
2007 }
2008 } else {
2009 if (!(isset($aUrl['path']) && $aUrl['path'] == $secremoveimg)) {
2010 // parse_url did not lead to satisfying result
2011 $attvalue = $sQuote . SM_PATH . 'images/blank.png' . $sQuote;
2012 }
2013 }
2014 }
2015 }
2016}
2017
2018/**
8bd0068d 2019 * This function edits the style definition to make them friendly and
2020 * usable in SquirrelMail.
2021 *
2022 * @param $message the message object
2023 * @param $id the message id
2024 * @param $content a string with whatever is between <style> and </style>
2025 * @param $mailbox the message mailbox
2026 * @return a string with edited content.
2027 */
e60a299a 2028function sq_fixstyle($body, $pos, $message, $id, $mailbox){
b583c3e8 2029 $me = 'sq_fixstyle';
7e2ff844 2030 // workaround for </style> in between comments
2031 $iCurrentPos = $pos;
2032 $content = '';
2033 $sToken = '';
2034 $bSucces = false;
2035 $bEndTag = false;
2036 for ($i=$pos,$iCount=strlen($body);$i<$iCount;++$i) {
2037 $char = $body{$i};
2038 switch ($char) {
2039 case '<':
10c0caeb 2040 $sToken = $char;
7e2ff844 2041 break;
2042 case '/':
2043 if ($sToken == '<') {
2044 $sToken .= $char;
2045 $bEndTag = true;
2046 } else {
2047 $content .= $char;
2048 }
2049 break;
2050 case '>':
2051 if ($bEndTag) {
2052 $sToken .= $char;
2053 if (preg_match('/\<\/\s*style\s*\>/i',$sToken,$aMatch)) {
2054 $newpos = $i + 1;
2055 $bSucces = true;
2056 break 2;
2057 } else {
2058 $content .= $sToken;
2059 }
2060 $bEndTag = false;
2061 } else {
2062 $content .= $char;
2063 }
2064 break;
2065 case '!':
2066 if ($sToken == '<') {
2067 // possible comment
2068 if (isset($body{$i+2}) && substr($body,$i,3) == '!--') {
2069 $i = strpos($body,'-->',$i+3);
2b665f28 2070 if ($i === false) { // no end comment
2071 $i = strlen($body);
2072 }
7e2ff844 2073 $sToken = '';
2074 }
2075 } else {
2076 $content .= $char;
2077 }
2078 break;
2079 default:
2080 if ($bEndTag) {
2081 $sToken .= $char;
2082 } else {
2083 $content .= $char;
2084 }
2085 break;
2086 }
2087 }
2088 if ($bSucces == FALSE){
7d06541f 2089 return array(FALSE, strlen($body));
2090 }
7e2ff844 2091
2b665f28 2092
2093
691a2d25 2094 /**
9ebace19 2095 * First look for general BODY style declaration, which would be
2096 * like so:
2097 * body {background: blah-blah}
2098 * and change it to .bodyclass so we can just assign it to a <div>
2099 */
691a2d25 2100 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
3d8371be 2101 $secremoveimg = '../images/' . _("sec_remove_eng.png");
691a2d25 2102 /**
0493ed11 2103 * Fix url('blah') declarations.
2104 */
2105 // $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
2106 // "url(\\1$secremoveimg\\2)", $content);
2b665f28 2107
567dc524 2108 // first check for 8bit sequences and disallowed control characters
2109 if (preg_match('/[\16-\37\200-\377]+/',$content)) {
2110 $content = '<!-- style block removed by html filter due to presence of 8bit characters -->';
2111 return array($content, $newpos);
2112 }
2113
2b665f28 2114 // IE Sucks hard. We have a special function for it.
2115 sq_fixIE_idiocy($content);
2116
2117 // remove @import line
2118 $content = preg_replace("/^\s*(@import.*)$/mi","\n<!-- @import rules forbidden -->\n",$content);
2119
5b4884be 2120 // translate ur\l and variations (IE parses that)
2b665f28 2121 // TODO check if the sq_fixIE_idiocy function already handles this.
5b4884be 2122 $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
567dc524 2123 preg_match_all("/url\s*\((.+)\)/si",$content,$aMatch);
2124 if (count($aMatch)) {
2125 $aValue = $aReplace = array();
2126 foreach($aMatch[1] as $sMatch) {
2127 // url value
2128 $urlvalue = $sMatch;
2129 sq_fix_url('style',$urlvalue, $message, $id, $mailbox,"'");
2130 $aValue[] = $sMatch;
2131 $aReplace[] = $urlvalue;
0493ed11 2132 }
567dc524 2133 $content = str_replace($aValue,$aReplace,$content);
691a2d25 2134 }
567dc524 2135
9ebace19 2136 /**
2137 * Remove any backslashes, entities, and extraneous whitespace.
2138 */
0493ed11 2139 $contentTemp = $content;
2140 sq_defang($contentTemp);
2141 sq_unspace($contentTemp);
a3daaaf3 2142
691a2d25 2143 /**
8bd0068d 2144 * Fix stupid css declarations which lead to vulnerabilities
2145 * in IE.
2146 */
5db90261 2147 $match = Array('/\/\*.*\*\//',
2148 '/expression/i',
0493ed11 2149 '/behaviou*r/i',
2150 '/binding/i',
2b665f28 2151 '/include-source/i',
2152 '/javascript/i',
2153 '/script/i');
2154 $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy');
0493ed11 2155 $contentNew = preg_replace($match, $replace, $contentTemp);
2156 if ($contentNew !== $contentTemp) {
2157 // insecure css declarations are used. From now on we don't care
2158 // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
2159 $content = $contentNew;
2160 }
7d06541f 2161 return array($content, $newpos);
691a2d25 2162}
a3daaaf3 2163
0493ed11 2164
691a2d25 2165/**
8bd0068d 2166 * This function converts cid: url's into the ones that can be viewed in
2167 * the browser.
2168 *
2169 * @param $message the message object
2170 * @param $id the message id
2171 * @param $cidurl the cid: url.
2172 * @param $mailbox the message mailbox
2173 * @return a string with a http-friendly url
2174 */
b3af12ef 2175function sq_cid2http($message, $id, $cidurl, $mailbox){
691a2d25 2176 /**
8bd0068d 2177 * Get rid of quotes.
2178 */
691a2d25 2179 $quotchar = substr($cidurl, 0, 1);
2dd879b8 2180 if ($quotchar == '"' || $quotchar == "'"){
2181 $cidurl = str_replace($quotchar, "", $cidurl);
2182 } else {
2183 $quotchar = '';
2184 }
691a2d25 2185 $cidurl = substr(trim($cidurl), 4);
0493ed11 2186
2187 $match_str = '/\{.*?\}\//';
2188 $str_rep = '';
2189 $cidurl = preg_replace($match_str, $str_rep, $cidurl);
2190
e5e9381a 2191 $linkurl = find_ent_id($cidurl, $message);
9ebace19 2192 /* in case of non-safe cid links $httpurl should be replaced by a sort of
2193 unsafe link image */
e5e9381a 2194 $httpurl = '';
c8f5f606 2195
8bd0068d 2196 /**
2197 * This is part of a fix for Outlook Express 6.x generating
2198 * cid URLs without creating content-id headers. These images are
2199 * not part of the multipart/related html mail. The html contains
2200 * <img src="cid:{some_id}/image_filename.ext"> references to
2201 * attached images with as goal to render them inline although
2202 * the attachment disposition property is not inline.
2203 */
c8f5f606 2204
2205 if (empty($linkurl)) {
2206 if (preg_match('/{.*}\//', $cidurl)) {
2207 $cidurl = preg_replace('/{.*}\//','', $cidurl);
2208 if (!empty($cidurl)) {
2209 $linkurl = find_ent_id($cidurl, $message);
2210 }
2211 }
2212 }
f8a1ed5a 2213
c8f5f606 2214 if (!empty($linkurl)) {
fa26ab45 2215 $httpurl = $quotchar . sqm_baseuri() . 'src/download.php?absolute_dl=true&amp;' .
8bd0068d 2216 "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
2217 '&amp;ent_id=' . $linkurl . $quotchar;
c8f5f606 2218 } else {
2219 /**
2220 * If we couldn't generate a proper img url, drop in a blank image
2221 * instead of sending back empty, otherwise it causes unusual behaviour
2222 */
bc017c1d 2223 $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
e5e9381a 2224 }
f8a1ed5a 2225
691a2d25 2226 return $httpurl;
2227}
2228
2229/**
8bd0068d 2230 * This function changes the <body> tag into a <div> tag since we
2231 * can't really have a body-within-body.
2232 *
2233 * @param $attary an array of attributes and values of <body>
2234 * @param $mailbox mailbox we're currently reading (for cid2http)
2235 * @param $message current message (for cid2http)
2236 * @param $id current message id (for cid2http)
2237 * @return a modified array of attributes to be set for <div>
2238 */
2dd879b8 2239function sq_body2div($attary, $mailbox, $message, $id){
b583c3e8 2240 $me = 'sq_body2div';
3d8371be 2241 $divattary = Array('class' => "'bodyclass'");
b583c3e8 2242 $text = '#000000';
c189a963 2243 $has_bgc_stl = $has_txt_stl = false;
b583c3e8 2244 $styledef = '';
691a2d25 2245 if (is_array($attary) && sizeof($attary) > 0){
2246 foreach ($attary as $attname=>$attvalue){
2247 $quotchar = substr($attvalue, 0, 1);
2248 $attvalue = str_replace($quotchar, "", $attvalue);
2249 switch ($attname){
3d8371be 2250 case 'background':
8bd0068d 2251 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
3d8371be 2252 $styledef .= "background-image: url('$attvalue'); ";
2253 break;
2254 case 'bgcolor':
c189a963 2255 $has_bgc_stl = true;
3d8371be 2256 $styledef .= "background-color: $attvalue; ";
2257 break;
2258 case 'text':
c189a963 2259 $has_txt_stl = true;
3d8371be 2260 $styledef .= "color: $attvalue; ";
2261 break;
691a2d25 2262 }
a3daaaf3 2263 }
c189a963 2264 // Outlook defines a white bgcolor and no text color. This can lead to
2265 // white text on a white bg with certain themes.
2266 if ($has_bgc_stl && !$has_txt_stl) {
2267 $styledef .= "color: $text; ";
2268 }
691a2d25 2269 if (strlen($styledef) > 0){
2270 $divattary{"style"} = "\"$styledef\"";
2271 }
2272 }
2273 return $divattary;
2274}
a3daaaf3 2275
691a2d25 2276/**
8bd0068d 2277 * This is the main function and the one you should actually be calling.
2278 * There are several variables you should be aware of an which need
2279 * special description.
2280 *
2281 * Since the description is quite lengthy, see it here:
2282 * http://linux.duke.edu/projects/mini/htmlfilter/
2283 *
2284 * @param $body the string with HTML you wish to filter
2285 * @param $tag_list see description above
2286 * @param $rm_tags_with_content see description above
2287 * @param $self_closing_tags see description above
2288 * @param $force_tag_closing see description above
2289 * @param $rm_attnames see description above
2290 * @param $bad_attvals see description above
2291 * @param $add_attr_to_tag see description above
2292 * @param $message message object
2293 * @param $id message id
2294 * @return sanitized html safe to show on your pages.
2295 */
da2415c1 2296function sq_sanitize($body,
8bd0068d 2297 $tag_list,
2298 $rm_tags_with_content,
2299 $self_closing_tags,
2300 $force_tag_closing,
2301 $rm_attnames,
2302 $bad_attvals,
2303 $add_attr_to_tag,
2304 $message,
2305 $id,
2306 $mailbox
2307 ){
b583c3e8 2308 $me = 'sq_sanitize';
7d06541f 2309 $rm_tags = array_shift($tag_list);
691a2d25 2310 /**
8bd0068d 2311 * Normalize rm_tags and rm_tags_with_content.
2312 */
7d06541f 2313 @array_walk($tag_list, 'sq_casenormalize');
691a2d25 2314 @array_walk($rm_tags_with_content, 'sq_casenormalize');
2315 @array_walk($self_closing_tags, 'sq_casenormalize');
2316 /**
8bd0068d 2317 * See if tag_list is of tags to remove or tags to allow.
2318 * false means remove these tags
2319 * true means allow these tags
2320 */
691a2d25 2321 $curpos = 0;
2322 $open_tags = Array();
2dd879b8 2323 $trusted = "\n<!-- begin sanitized html -->\n";
691a2d25 2324 $skip_content = false;
bb8d0799 2325 /**
8bd0068d 2326 * Take care of netscape's stupid javascript entities like
2327 * &{alert('boo')};
2328 */
bb8d0799 2329 $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
691a2d25 2330
7d06541f 2331 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
691a2d25 2332 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
2333 $free_content = substr($body, $curpos, $lt-$curpos);
2334 /**
8bd0068d 2335 * Take care of <style>
2336 */
7d06541f 2337 if ($tagname == "style" && $tagtype == 1){
da2415c1 2338 list($free_content, $curpos) =
e60a299a 2339 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
7d06541f 2340 if ($free_content != FALSE){
2341 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2342 $trusted .= $free_content;
2343 $trusted .= sq_tagprint($tagname, false, 2);
2344 }
2345 continue;
691a2d25 2346 }
2347 if ($skip_content == false){
2348 $trusted .= $free_content;
691a2d25 2349 }
2350 if ($tagname != FALSE){
2351 if ($tagtype == 2){
2352 if ($skip_content == $tagname){
2353 /**
8bd0068d 2354 * Got to the end of tag we needed to remove.
2355 */
691a2d25 2356 $tagname = false;
2357 $skip_content = false;
2358 } else {
2359 if ($skip_content == false){
c828931c 2360 if ($tagname == "body"){
2361 $tagname = "div";
2dd879b8 2362 }
da2415c1 2363 if (isset($open_tags{$tagname}) &&
8bd0068d 2364 $open_tags{$tagname} > 0){
2dd879b8 2365 $open_tags{$tagname}--;
691a2d25 2366 } else {
2dd879b8 2367 $tagname = false;
691a2d25 2368 }
691a2d25 2369 }
2370 }
2371 } else {
2372 /**
8bd0068d 2373 * $rm_tags_with_content
2374 */
691a2d25 2375 if ($skip_content == false){
2376 /**
8bd0068d 2377 * See if this is a self-closing type and change
2378 * tagtype appropriately.
2379 */
691a2d25 2380 if ($tagtype == 1
8bd0068d 2381 && in_array($tagname, $self_closing_tags)){
2dd879b8 2382 $tagtype = 3;
691a2d25 2383 }
2384 /**
8bd0068d 2385 * See if we should skip this tag and any content
2386 * inside it.
2387 */
691a2d25 2388 if ($tagtype == 1 &&
8bd0068d 2389 in_array($tagname, $rm_tags_with_content)){
691a2d25 2390 $skip_content = $tagname;
2391 } else {
da2415c1 2392 if (($rm_tags == false
8bd0068d 2393 && in_array($tagname, $tag_list)) ||
2394 ($rm_tags == true &&
2395 !in_array($tagname, $tag_list))){
691a2d25 2396 $tagname = false;
2397 } else {
2dd879b8 2398 /**
8bd0068d 2399 * Convert body into div.
2400 */
2dd879b8 2401 if ($tagname == "body"){
2402 $tagname = "div";
da2415c1 2403 $attary = sq_body2div($attary, $mailbox,
8bd0068d 2404 $message, $id);
2dd879b8 2405 }
691a2d25 2406 if ($tagtype == 1){
2407 if (isset($open_tags{$tagname})){
2408 $open_tags{$tagname}++;
2409 } else {
2410 $open_tags{$tagname}=1;
2411 }
2412 }
2413 /**
8bd0068d 2414 * This is where we run other checks.
2415 */
691a2d25 2416 if (is_array($attary) && sizeof($attary) > 0){
2417 $attary = sq_fixatts($tagname,
8bd0068d 2418 $attary,
2419 $rm_attnames,
2420 $bad_attvals,
2421 $add_attr_to_tag,
2422 $message,
2423 $id,
2424 $mailbox
2425 );
691a2d25 2426 }
2427 }
2428 }
691a2d25 2429 }
2430 }
2431 if ($tagname != false && $skip_content == false){
2432 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2433 }
691a2d25 2434 }
2435 $curpos = $gt+1;
a3daaaf3 2436 }
691a2d25 2437 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2438 if ($force_tag_closing == true){
2439 foreach ($open_tags as $tagname=>$opentimes){
2440 while ($opentimes > 0){
2441 $trusted .= '</' . $tagname . '>';
2442 $opentimes--;
2443 }
2444 }
2445 $trusted .= "\n";
2446 }
2447 $trusted .= "<!-- end sanitized html -->\n";
2448 return $trusted;
2449}
451f74a2 2450
691a2d25 2451/**
8bd0068d 2452 * This is a wrapper function to call html sanitizing routines.
2453 *
2454 * @param $body the body of the message
2455 * @param $id the id of the message
c189a963 2456
2457 * @param $message
2458 * @param $mailbox
2459 * @param boolean $take_mailto_links When TRUE, converts mailto: links
2460 * into internal SM compose links
2461 * (optional; default = TRUE)
8bd0068d 2462 * @return a string with html safe to display in the browser.
2463 */
c189a963 2464function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2465
202bcbcc 2466 // require_once(SM_PATH . 'functions/url_parser.php'); // for $MailTo_PReg_Match
c189a963 2467
691a2d25 2468 global $attachment_common_show_images, $view_unsafe_images,
8bd0068d 2469 $has_unsafe_images;
691a2d25 2470 /**
8bd0068d 2471 * Don't display attached images in HTML mode.
2b665f28 2472 *
d0187bd6 2473 * SB: why?
8bd0068d 2474 */
691a2d25 2475 $attachment_common_show_images = false;
2476 $tag_list = Array(
8bd0068d 2477 false,
2478 "object",
2479 "meta",
2480 "html",
2481 "head",
2482 "base",
2483 "link",
2484 "frame",
2485 "iframe",
2486 "plaintext",
2487 "marquee"
2488 );
691a2d25 2489
2490 $rm_tags_with_content = Array(
8bd0068d 2491 "script",
2492 "applet",
2493 "embed",
2494 "title",
2495 "frameset",
0493ed11 2496 "xmp",
8bd0068d 2497 "xml"
2498 );
691a2d25 2499
2500 $self_closing_tags = Array(
8bd0068d 2501 "img",
2502 "br",
2503 "hr",
2504 "input",
2505 "outbind"
2506 );
691a2d25 2507
2dd879b8 2508 $force_tag_closing = true;
691a2d25 2509
2510 $rm_attnames = Array(
8bd0068d 2511 "/.*/" =>
2512 Array(
2513 "/target/i",
2514 "/^on.*/i",
2515 "/^dynsrc/i",
2516 "/^data.*/i",
2517 "/^lowsrc.*/i"
2518 )
2519 );
691a2d25 2520
2521 $secremoveimg = "../images/" . _("sec_remove_eng.png");
2522 $bad_attvals = Array(
8bd0068d 2523 "/.*/" =>
691a2d25 2524 Array(
0a6ec9b5 2525 "/^src|background/i" =>
8bd0068d 2526 Array(
691a2d25 2527 Array(
8bd0068d 2528 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2529 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2530 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
691a2d25 2531 ),
8bd0068d 2532 Array(
2533 "\\1$secremoveimg\\2",
2534 "\\1$secremoveimg\\2",
2535 "\\1$secremoveimg\\2",
8bd0068d 2536 )
2537 ),
0a6ec9b5 2538 "/^href|action/i" =>
8bd0068d 2539 Array(
0a6ec9b5 2540 Array(
8bd0068d 2541 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2542 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2543 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
0a6ec9b5 2544 ),
691a2d25 2545 Array(
8bd0068d 2546 "\\1#\\1",
2547 "\\1#\\1",
8bd0068d 2548 "\\1#\\1"
02474e43 2549 )
8bd0068d 2550 ),
2551 "/^style/i" =>
2552 Array(
2553 Array(
5db90261 2554 "/\/\*.*\*\//",
8bd0068d 2555 "/expression/i",
2556 "/binding/i",
2557 "/behaviou*r/i",
2558 "/include-source/i",
2559 "/position\s*:\s*absolute/i",
1d935bc2 2560 "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
8bd0068d 2561 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2562 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2563 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2564 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2565 ),
2566 Array(
5db90261 2567 "",
8bd0068d 2568 "idiocy",
2569 "idiocy",
2570 "idiocy",
2571 "idiocy",
567dc524 2572 "idiocy",
5b4884be 2573 "url",
8bd0068d 2574 "url(\\1#\\1)",
2575 "url(\\1#\\1)",
2576 "url(\\1#\\1)",
8bd0068d 2577 "\\1:url(\\2#\\3)"
2578 )
691a2d25 2579 )
8bd0068d 2580 )
691a2d25 2581 );
9ebace19 2582
d31f73f1 2583 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
2584 // images off by default.
afbf184c 2585 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET, FALSE);
9ebace19 2586
691a2d25 2587 if (!$view_unsafe_images){
2588 /**
8bd0068d 2589 * Remove any references to http/https if view_unsafe_images set
2590 * to false.
2591 */
02474e43 2592 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
8bd0068d 2593 '/^([\'\"])\s*https*:.*([\'\"])/si');
02474e43 2594 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
8bd0068d 2595 "\\1$secremoveimg\\1");
02474e43 2596 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
7ef0b415 2597 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
02474e43 2598 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
8bd0068d 2599 "url(\\1$secremoveimg\\1)");
691a2d25 2600 }
451f74a2 2601
691a2d25 2602 $add_attr_to_tag = Array(
8bd0068d 2603 "/^a$/i" =>
c25f2fbb 2604 Array('target'=>'"_blank"',
02474e43 2605 'title'=>'"'._("This external link will open in a new window").'"'
8bd0068d 2606 )
2607 );
da2415c1 2608 $trusted = sq_sanitize($body,
8bd0068d 2609 $tag_list,
2610 $rm_tags_with_content,
2611 $self_closing_tags,
2612 $force_tag_closing,
2613 $rm_attnames,
2614 $bad_attvals,
2615 $add_attr_to_tag,
2616 $message,
2617 $id,
2618 $mailbox
2619 );
567dc524 2620 if (strpos($trusted,$secremoveimg)){
691a2d25 2621 $has_unsafe_images = true;
da2415c1 2622 }
c189a963 2623
2624 // we want to parse mailto's in HTML output, change to SM compose links
2625 // this is a modified version of code from url_parser.php... but Marc is
2626 // right: we need a better filtering implementation; adding this randomly
2627 // here is not a great solution
2628 //
2629 if ($take_mailto_links) {
2630 // parseUrl($trusted); // this even parses URLs inside of tags... too aggressive
2631 global $MailTo_PReg_Match;
202bcbcc 2632 $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1) ;
c189a963 2633 if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2634 foreach ($regs[0] as $i => $mailto_before) {
2635 $mailto_params = $regs[10][$i];
2636 // get rid of any tailing quote since we have to add send_to to the end
2637 //
2638 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2639 $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2640 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2641 $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2642
2643 if ($regs[1][$i]) { //if there is an email addr before '?', we need to merge it with the params
2644 $to = 'to=' . $regs[1][$i];
2645 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
2646 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2647 else {
2648 if ($mailto_params) //already some params, append to them
2649 $mailto_params .= '&amp;' . $to;
2650 else
2651 $mailto_params .= '?' . $to;
2652 }
2653 }
2654
2655 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2656
2657 // we'll already have target=_blank, no need to allow comp_in_new
2658 // here (which would be a lot more work anyway)
2659 //
2660 global $compose_new_win;
2661 $temp_comp_in_new = $compose_new_win;
2662 $compose_new_win = 0;
2663 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2664 $compose_new_win = $temp_comp_in_new;
2665
2666 // remove <a href=" and anything after the next quote (we only
2667 // need the uri, not the link HTML) in compose uri
2668 //
2669 $comp_uri = substr($comp_uri, 9);
2670 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2671 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2672 }
2673 }
2674 }
2675
691a2d25 2676 return $trusted;
451f74a2 2677}
a4a70693 2678
da2415c1 2679/**
8bd0068d 2680 * function SendDownloadHeaders - send file to the browser
2681 *
2682 * Original Source: SM core src/download.php
2683 * moved here to make it available to other code, and separate
2684 * front end from back end functionality.
2685 *
2686 * @param string $type0 first half of mime type
2687 * @param string $type1 second half of mime type
2688 * @param string $filename filename to tell the browser for downloaded file
2689 * @param boolean $force whether to force the download dialog to pop
2690 * @param optional integer $filesize send the Content-Header and length to the browser
2691 * @return void
2692 */
02474e43 2693function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2694 global $languages, $squirrelmail_language;
cfffd60b 2695 $isIE = $isIE6plus = false;
02474e43 2696
2697 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2698
2699 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
8bd0068d 2700 strstr($HTTP_USER_AGENT, 'Opera') === false) {
cfffd60b 2701 $isIE = true;
02474e43 2702 }
2703
cfffd60b 2704 if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2705 ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2706 $isIE6plus = true;
02474e43 2707 }
2708
2709 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 2710 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
02474e43 2711 $filename =
8bd0068d 2712 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
02474e43 2713 } else {
2714 $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
2715 }
2716
2717 // A Pox on Microsoft and it's Internet Explorer!
2718 //
2719 // IE has lots of bugs with file downloads.
2720 // It also has problems with SSL. Both of these cause problems
2721 // for us in this function.
2722 //
2723 // See this article on Cache Control headers and SSL
2724 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2725 //
2726 // The best thing you can do for IE is to upgrade to the latest
2727 // version
2728 //set all the Cache Control Headers for IE
2729 if ($isIE) {
2730 $filename=rawurlencode($filename);
2731 header ("Pragma: public");
8bd0068d 2732 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
02474e43 2733 header ("Cache-Control: post-check=0, pre-check=0", false);
8bd0068d 2734 header ("Cache-Control: private");
02474e43 2735
2736 //set the inline header for IE, we'll add the attachment header later if we need it
2737 header ("Content-Disposition: inline; filename=$filename");
2738 }
2739
2740 if (!$force) {
2741 // Try to show in browser window
2742 header ("Content-Disposition: inline; filename=\"$filename\"");
2743 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2744 } else {
2745 // Try to pop up the "save as" box
2746
2747 // IE makes this hard. It pops up 2 save boxes, or none.
2748 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2749 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2750 // But, according to Microsoft, it is "RFC compliant but doesn't
2751 // take into account some deviations that allowed within the
2752 // specification." Doesn't that mean RFC non-compliant?
2753 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2754
2755 // all browsers need the application/octet-stream header for this
2756 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2757
2758 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2759 // Do not have quotes around filename, but that applied to
2760 // "attachment"... does it apply to inline too?
2761 header ("Content-Disposition: attachment; filename=\"$filename\"");
2762
cfffd60b 2763 if ($isIE && !$isIE6plus) {
02474e43 2764 // This combination seems to work mostly. IE 5.5 SP 1 has
2765 // known issues (see the Microsoft Knowledge Base)
2766
2767 // This works for most types, but doesn't work with Word files
2768 header ("Content-Type: application/download; name=\"$filename\"");
7e2ff844 2769 header ("Content-Type: application/force-download; name=\"$filename\"");
02474e43 2770 // These are spares, just in case. :-)
2771 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2772 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2773 //header("Content-Type: application/octet-stream; name=\"$filename\"");
7e2ff844 2774 } else if ($isIE) {
2775 // This is to prevent IE for MIME sniffing and auto open a file in IE
2776 header ("Content-Type: application/force-download; name=\"$filename\"");
02474e43 2777 } else {
2778 // another application/octet-stream forces download for Netscape
2779 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2780 }
2781 }
2782
2783 //send the content-length header if the calling function provides it
2784 if ($filesize > 0) {
2785 header("Content-Length: $filesize");
2786 }
07c49f57 2787
8d863f64 2788} // end fn SendDownloadHeaders