projects / squirrelmail.git / blame
? search:
summary | shortlog | log | commit | commitdiff | tree
blob | blame (incremental) | history | HEAD
Allow choice of line endings when downloading or decoding strings (bodies)
[squirrelmail.git] / functions / mime.php
CommitLineData
59177427 1<?php
2ba13803 2
35586184 3/**
8bd0068d 4 * mime.php
5 *
8bd0068d 6 * This contains the functions necessary to detect and decode MIME
7 * messages.
8 *
4b5049de 9 * @copyright &copy; 1999-2007 The SquirrelMail Project Team
4b4abf93 10 * @license http://opensource.org/licenses/gpl-license.php GNU Public License
8bd0068d 11 * @version $Id$
12 * @package squirrelmail
13 */
b74ba498 14
202bcbcc 15/**
16 * dependency information
17 functions dependency
18 mime_structure
19 class/mime/Message.class.php
20 Message::parseStructure
21 functions/page_header.php
22 displayPageHeader
23 functions/display_messages.php
24 plain_error_message
25 mime_fetch_body
26 functions/imap_general.php
27 sqimap_run_command
28 mime_print_body_lines
29
30
31
32functions/imap.php
33functions/attachment_common.php
34functions/display_messages.php
35
36magicHtml => url_parser
37translateText => url_parser
38
39*/
40
8beafbbc 41
7c7b74b3 42/* -------------------------------------------------------------------------- */
43/* MIME DECODING */
44/* -------------------------------------------------------------------------- */
b74ba498 45
d6c32258 46/**
8bd0068d 47 * Get the MIME structure
48 *
49 * This function gets the structure of a message and stores it in the "message" class.
50 * It will return this object for use with all relevant header information and
51 * fully parsed into the standard "message" object format.
52 */
a4a70693 53function mime_structure ($bodystructure, $flags=array()) {
c9d78ab4 54
3d8371be 55 /* Isolate the body structure and remove beginning and end parenthesis. */
a4a70693 56 $read = trim(substr ($bodystructure, strpos(strtolower($bodystructure), 'bodystructure') + 13));
451f74a2 57 $read = trim(substr ($read, 0, -1));
22efa9fb 58 $i = 0;
59 $msg = Message::parseStructure($read,$i);
2b665f28 60
9de42168 61 if (!is_object($msg)) {
3d8371be 62 global $color, $mailbox;
a48eba8f 63 displayPageHeader( $color, $mailbox );
5e8de8b6 64 $errormessage = _("SquirrelMail could not decode the bodystructure of the message");
8bd0068d 65 $errormessage .= '<br />'._("The bodystructure provided by your IMAP server:").'<br /><br />';
472e7acb 66 $errormessage .= '<pre>' . htmlspecialchars($read) . '</pre>';
ce8c6f42 67 plain_error_message( $errormessage );
3d8371be 68 echo '</body></html>';
9de42168 69 exit;
70 }
a4a70693 71 if (count($flags)) {
7a9e9c89 72 foreach ($flags as $flag) {
e1115979 73//FIXME: please document why it is we have to check the first char of the flag but we then go ahead and do a full string comparison anyway. Is this a speed enhancement? If not, let's keep it simple and just compare the full string and forget the switch block.
7a9e9c89 74 $char = strtoupper($flag{1});
75 switch ($char) {
3d8371be 76 case 'S':
77 if (strtolower($flag) == '\\seen') {
78 $msg->is_seen = true;
79 }
80 break;
81 case 'A':
82 if (strtolower($flag) == '\\answered') {
83 $msg->is_answered = true;
84 }
85 break;
86 case 'D':
87 if (strtolower($flag) == '\\deleted') {
88 $msg->is_deleted = true;
89 }
90 break;
91 case 'F':
92 if (strtolower($flag) == '\\flagged') {
93 $msg->is_flagged = true;
94 }
95 break;
96 case 'M':
97 if (strtolower($flag) == '$mdnsent') {
98 $msg->is_mdnsent = true;
99 }
100 break;
101 default:
102 break;
7a9e9c89 103 }
104 }
451f74a2 105 }
7a9e9c89 106 // listEntities($msg);
3d8371be 107 return $msg;
451f74a2 108}
b74ba498 109
22efa9fb 110
111
3d8371be 112/* This starts the parsing of a particular structure. It is called recursively,
8bd0068d 113 * so it can be passed different structures. It returns an object of type
114 * $message.
115 * First, it checks to see if it is a multipart message. If it is, then it
116 * handles that as it sees is necessary. If it is just a regular entity,
117 * then it parses it and adds the necessary header information (by calling out
118 * to mime_get_elements()
119 */
451f74a2 120
4d592352 121function mime_fetch_body($imap_stream, $id, $ent_id=1, $fetch_size=0) {
3d8371be 122 /* Do a bit of error correction. If we couldn't find the entity id, just guess
8bd0068d 123 * that it is the first one. That is usually the case anyway.
124 */
7c7b74b3 125
09a4bde3 126 if (!$ent_id) {
08b7f7cc 127 $cmd = "FETCH $id BODY[]";
1035e159 128 } else {
08b7f7cc 129 $cmd = "FETCH $id BODY[$ent_id]";
09a4bde3 130 }
3d8371be 131
4d592352 132 if ($fetch_size!=0) $cmd .= "<0.$fetch_size>";
da2415c1 133
6201339c 134 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, TRUE);
77b88425 135 do {
3d8371be 136 $topline = trim(array_shift($data));
137 } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
a4a70693 138
451f74a2 139 $wholemessage = implode('', $data);
140 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
3d8371be 141 $ret = substr($wholemessage, 0, $regs[1]);
142 /* There is some information in the content info header that could be important
8bd0068d 143 * in order to parse html messages. Let's get them here.
144 */
0600bdf1 145// if ($ret{0} == '<') {
6201339c 146// $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, TRUE);
0600bdf1 147// }
451f74a2 148 } else if (ereg('"([^"]*)"', $topline, $regs)) {
149 $ret = $regs[1];
06bcb9c3 150 } else if ((stristr($topline, 'nil') !== false) && (empty($wholemessage))) {
151 $ret = $wholemessage;
451f74a2 152 } else {
153 global $where, $what, $mailbox, $passed_id, $startMessage;
3d8371be 154 $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
451f74a2 155 if (isset($where) && isset($what)) {
3d8371be 156 $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
a3daaaf3 157 } else {
3d8371be 158 $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
451f74a2 159 }
e5ea9327 160 $par .= '&amp;response=' . urlencode($response) .
8bd0068d 161 '&amp;message=' . urlencode($message) .
162 '&amp;topline=' . urlencode($topline);
a019eeb8 163
8bd0068d 164 echo '<tt><br />' .
02474e43 165 '<table width="80%"><tr>' .
166 '<tr><td colspan="2">' .
167 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
168 '</td></tr>' .
169 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
170 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
171 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
172 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
173 "</table><br /></tt></font><hr />";
346817d4 174
6201339c 175 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, TRUE);
451f74a2 176 array_shift($data);
177 $wholemessage = implode('', $data);
a019eeb8 178
346817d4 179 $ret = $wholemessage;
a3daaaf3 180 }
3d8371be 181 return $ret;
451f74a2 182}
d4467150 183
fa26ab45 184function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding, $rStream='php://stdout', $force_crlf='') {
1035e159 185
3d8371be 186 /* Don't kill the connection if the browser is over a dialup
8bd0068d 187 * and it would take over 30 seconds to download it.
188 * Don't call set_time_limit in safe mode.
189 */
b7206e1d 190
3d8371be 191 if (!ini_get('safe_mode')) {
b7206e1d 192 set_time_limit(0);
193 }
7c7b74b3 194 /* in case of base64 encoded attachments, do not buffer them.
8bd0068d 195 Instead, echo the decoded attachment directly to screen */
7c7b74b3 196 if (strtolower($encoding) == 'base64') {
197 if (!$ent_id) {
7591f143 198 $query = "FETCH $id BODY[]";
7c7b74b3 199 } else {
7591f143 200 $query = "FETCH $id BODY[$ent_id]";
7c7b74b3 201 }
7591f143 202 sqimap_run_command($imap_stream,$query,true,$response,$message,TRUE,'sqimap_base64_decode',$rStream,true);
1d142b8d 203 } else {
7591f143 204 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
205 if (is_resource($rStream)) {
fa26ab45 206 fputs($rStream,decodeBody($body, $encoding, $force_crlf));
7591f143 207 } else {
fa26ab45 208 echo decodeBody($body, $encoding, $force_crlf);
7591f143 209 }
1d142b8d 210 }
346817d4 211
da2415c1 212 /*
8bd0068d 213 TODO, use the same method for quoted printable.
214 However, I assume that quoted printable attachments aren't that large
215 so the performancegain / memory usage drop will be minimal.
216 If we decide to add that then we need to adapt sqimap_fread because
217 we need to split te result on \n and fread doesn't stop at \n. That
218 means we also should provide $results from sqimap_fread (by ref) to
219 te function and set $no_return to false. The $filter function for
220 quoted printable should handle unsetting of $results.
221 */
da2415c1 222 /*
8bd0068d 223 TODO 2: find out how we write to the output stream php://stdout. fwrite
224 doesn't work because 'php://stdout isn't a stream.
225 */
7c7b74b3 226
5d9c6f73 227 return;
451f74a2 228}
beb9e459 229
451f74a2 230/* -[ END MIME DECODING ]----------------------------------------------------------- */
d4467150 231
3d8371be 232/* This is here for debugging purposes. It will print out a list
8bd0068d 233 * of all the entity IDs that are in the $message object.
234 */
451f74a2 235function listEntities ($message) {
3d8371be 236 if ($message) {
3c621ba1 237 echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br />';
3d8371be 238 for ($i = 0; isset($message->entities[$i]); $i++) {
239 echo "$i : ";
240 $msg = listEntities($message->entities[$i]);
241
242 if ($msg) {
243 echo "return: ";
244 return $msg;
245 }
246 }
a4a70693 247 }
451f74a2 248}
f0c4dc12 249
f792c641 250function getPriorityStr($priority) {
3d8371be 251 $priority_level = substr($priority,0,1);
252
253 switch($priority_level) {
254 /* Check for a higher then normal priority. */
255 case '1':
256 case '2':
257 $priority_string = _("High");
258 break;
259
260 /* Check for a lower then normal priority. */
261 case '4':
262 case '5':
263 $priority_string = _("Low");
264 break;
265
266 /* Check for a normal priority. */
267 case '3':
268 default:
269 $priority_level = '3';
270 $priority_string = _("Normal");
271 break;
272
273 }
274 return $priority_string;
f792c641 275}
276
451f74a2 277/* returns a $message object for a particular entity id */
278function getEntity ($message, $ent_id) {
a4a70693 279 return $message->getEntity($ent_id);
451f74a2 280}
8beafbbc 281
3d8371be 282/* translateText
8bd0068d 283 * Extracted from strings.php 23/03/2002
284 */
da4c66e8 285
286function translateText(&$body, $wrap_at, $charset) {
3d8371be 287 global $where, $what; /* from searching */
288 global $color; /* color theme */
da4c66e8 289
202bcbcc 290 // require_once(SM_PATH . 'functions/url_parser.php');
da4c66e8 291
292 $body_ary = explode("\n", $body);
da4c66e8 293 for ($i=0; $i < count($body_ary); $i++) {
294 $line = $body_ary[$i];
295 if (strlen($line) - 2 >= $wrap_at) {
c7aff938 296 sqWordWrap($line, $wrap_at, $charset);
da4c66e8 297 }
298 $line = charset_decode($charset, $line);
299 $line = str_replace("\t", ' ', $line);
300
301 parseUrl ($line);
302
3d8371be 303 $quotes = 0;
da4c66e8 304 $pos = 0;
3d8371be 305 $j = strlen($line);
da4c66e8 306
3d8371be 307 while ($pos < $j) {
da4c66e8 308 if ($line[$pos] == ' ') {
3d8371be 309 $pos++;
da4c66e8 310 } else if (strpos($line, '&gt;', $pos) === $pos) {
311 $pos += 4;
3d8371be 312 $quotes++;
da4c66e8 313 } else {
314 break;
315 }
316 }
3d8371be 317
83c94382 318 if ($quotes % 2) {
d0814c02 319 $line = '<span class="quote1">' . $line . '</span>';
4c25967c 320 } elseif ($quotes) {
d0814c02 321 $line = '<span class="quote2">' . $line . '</span>';
da4c66e8 322 }
3d8371be 323
da4c66e8 324 $body_ary[$i] = $line;
325 }
326 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
327}
328
a2bfcbce 329/**
da1b55ad 330 * This returns a parsed string called $body. That string can then
331 * be displayed as the actual message in the HTML. It contains
332 * everything needed, including HTML Tags, Attachments at the
333 * bottom, etc.
f8a1ed5a 334 *
da1b55ad 335 * Since 1.2.0 function uses message_body hook.
336 * Till 1.3.0 function included output of formatAttachments().
337 *
338 * @param resource $imap_stream imap connection resource
339 * @param object $message squirrelmail message object
340 * @param array $color squirrelmail color theme array
341 * @param integer $wrap_at number of characters per line
342 * @param string $ent_num (since 1.3.0) message part id
343 * @param integer $id (since 1.3.0) message id
344 * @param string $mailbox (since 1.3.0) imap folder name
da1b55ad 345 * @return string html formated message text
346 */
a540f994 347function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
3d8371be 348 /* This if statement checks for the entity to show as the
8bd0068d 349 * primary message. To add more of them, just put them in the
350 * order that is their priority.
351 */
ce68b76b 352 global $startMessage, $languages, $squirrelmail_language,
40a34e57 353 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id,
14c85e39 354 $use_iframe, $iframe_height, $download_and_unsafe_link,
955bfc8f 355 $download_href, $unsafe_image_toggle_href, $unsafe_image_toggle_text,
551c7b53 356 $oTemplate, $nbsp;
2c25d36a 357
358 // workaround for not updated config.php
359 if (! isset($use_iframe)) $use_iframe = false;
77bfbd2e 360
d31f73f1 361 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
362 // images off by default.
afbf184c 363 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET, FALSE);
d03c24f4 364
cc34b00d 365 $body = '';
23bcec6f 366 $urlmailbox = urlencode($mailbox);
451f74a2 367 $body_message = getEntity($message, $ent_num);
368 if (($body_message->header->type0 == 'text') ||
8bd0068d 369 ($body_message->header->type0 == 'rfc822')) {
3d8371be 370 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
451f74a2 371 $body = decodeBody($body, $body_message->header->encoding);
e842b215 372
373 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 374 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
e842b215 375 if (mb_detect_encoding($body) != 'ASCII') {
33a55f5a 376 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
e842b215 377 }
378 }
d849b570 379
380 /* As of 1.5.2, $body is passed (and modified) by reference */
381 do_hook('message_body', $body);
23bcec6f 382
3d8371be 383 /* If there are other types that shouldn't be formatted, add
8bd0068d 384 * them here.
385 */
3d8371be 386
451f74a2 387 if ($body_message->header->type1 == 'html') {
3d8371be 388 if ($show_html_default <> 1) {
85015544 389 $entity_conv = array('&nbsp;' => ' ',
8bd0068d 390 '<p>' => "\n",
391 '<P>' => "\n",
392 '<br>' => "\n",
393 '<BR>' => "\n",
394 '<br />' => "\n",
395 '<BR />' => "\n",
396 '&gt;' => '>',
397 '&lt;' => '<');
85015544 398 $body = strtr($body, $entity_conv);
3d8371be 399 $body = strip_tags($body);
85015544 400 $body = trim($body);
401 translateText($body, $wrap_at,
8bd0068d 402 $body_message->header->getParameter('charset'));
a540f994 403 } elseif ($use_iframe) {
84410f31 404 /**
405 * If we don't add html message between iframe tags,
406 * we must detect unsafe images and modify $has_unsafe_images.
f8a1ed5a 407 */
758a7889 408 $html_body = magicHTML($body, $id, $message, $mailbox);
b6c52e61 409 // Convert character set in order to display html mails in different character set
410 $html_body = charset_decode($body_message->header->getParameter('charset'),$html_body,false,true);
84410f31 411
2c25d36a 412 // creating iframe url
413 $iframeurl=sqm_baseuri().'src/view_html.php?'
f8a1ed5a 414 . 'mailbox=' . $urlmailbox
2c25d36a 415 . '&amp;passed_id=' . $id
416 . '&amp;ent_id=' . $ent_num
417 . '&amp;view_unsafe_images=' . (int) $view_unsafe_images;
418
79d58d4c 419 global $oTemplate;
420 $oTemplate->assign('iframe_url', $iframeurl);
51554708 421 $oTemplate->assign('iframe_height', $iframe_height);
79d58d4c 422 $oTemplate->assign('html_body', $html_body);
2b665f28 423
79d58d4c 424 $body = $oTemplate->fetch('read_html_iframe.tpl');
a3daaaf3 425 } else {
2c25d36a 426 // old way of html rendering
b6c52e61 427 /**
758a7889 428 * convert character set. charset_decode does not remove html special chars
b6c52e61 429 * applied by magicHTML functions and does not sanitize them second time if
758a7889 430 * fourth argument is true.
431 */
567dc524 432 $charset = $body_message->header->getParameter('charset');
433 if (!empty($charset)) {
434 $body = charset_decode($charset,$body,false,true);
435 }
436 $body = magicHTML($body, $id, $message, $mailbox);
a3daaaf3 437 }
451f74a2 438 } else {
3d8371be 439 translateText($body, $wrap_at,
8bd0068d 440 $body_message->header->getParameter('charset'));
451f74a2 441 }
a2bfcbce 442
9ebace19 443 /*
444 * Previously the links for downloading and unsafe images were printed
445 * under the mail. By putting the links in a global variable we can
446 * print it in the toolbar where it belongs. Since the original code was
447 * in this place it's left here. It might be possible to move it to some
448 * other place if that makes sense. The possibility to do so has not
449 * been evaluated yet.
450 */
451
452 // Initialize the global variable to an empty string.
453 // FIXME: To have $download_and_unsafe_link as a global variable might not be needed since the use of separate variables ($download_href, $unsafe_image_toggle_href, and $unsafe_image_toggle_text) for the templates was introduced.
40a34e57 454 $download_and_unsafe_link = '';
455
9ebace19 456 // Prepare and build a link for downloading the mail.
83cf04bd 457 $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
8bd0068d 458 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
459 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
08b7f7cc 460 if (isset($passed_ent_id)) {
461 $link .= '&amp;passed_ent_id='.$passed_ent_id;
462 }
14c85e39 463 $download_href = SM_PATH . 'src/download.php?absolute_dl=true&amp;' . $link;
9ebace19 464
465 // Always add the link for downloading the mail as a file to the global
466 // variable.
955bfc8f 467 $download_and_unsafe_link .= "$nbsp|$nbsp"
468 . create_hyperlink($download_href, _("Download this as a file"));
9ebace19 469
470 // Find out the right text to use in the link depending on the
471 // circumstances. If the unsafe images are displayed the link should
472 // hide them, if they aren't displayed the link should only appear if
473 // the mail really contains unsafe images.
7aad7b77 474 if ($view_unsafe_images) {
23f617b8 475 $text = _("Hide Unsafe Images");
7aad7b77 476 } else {
08b7f7cc 477 if (isset($has_unsafe_images) && $has_unsafe_images) {
478 $link .= '&amp;view_unsafe_images=1';
479 $text = _("View Unsafe Images");
480 } else {
481 $text = '';
482 }
3d8371be 483 }
9ebace19 484
485 // Only create a link for unsafe images if there's need for one. If so:
486 // add it to the global variable.
83cf04bd 487 if($text != '') {
14c85e39 488 $unsafe_image_toggle_href = SM_PATH . 'src/read_body.php?'.$link;
489 $unsafe_image_toggle_text = $text;
955bfc8f 490 $download_and_unsafe_link .= "$nbsp|$nbsp"
491 . create_hyperlink($unsafe_image_toggle_href, $text);
83cf04bd 492 }
3d8371be 493 }
494 return $body;
451f74a2 495}
b74ba498 496
da1b55ad 497/**
a540f994 498 * Generate attachments array for passing to templates.
2b665f28 499 *
d67f519a 500 * @since 1.5.2
da1b55ad 501 * @param object $message SquirrelMail message object
502 * @param array $exclude_id message parts that are not attachments.
503 * @param string $mailbox mailbox name
504 * @param integer $id message id
da1b55ad 505 */
d67f519a 506function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
507 global $where, $what, $startMessage, $color, $passed_ent_id, $base_uri;
451f74a2 508
23bcec6f 509 $att_ar = $message->getAttachments($exclude_id);
23bcec6f 510 $urlMailbox = urlencode($mailbox);
511
d67f519a 512 $attachments = array();
23bcec6f 513 foreach ($att_ar as $att) {
fdc9d9b5 514 $ent = $att->entity_id;
2e25760a 515 $header = $att->header;
f0c4dc12 516 $type0 = strtolower($header->type0);
517 $type1 = strtolower($header->type1);
2e25760a 518 $name = '';
d0187bd6 519 $links = array();
21dab2dc 520 $links['download link']['text'] = _("Download");
202bcbcc 521 $links['download link']['href'] = $base_uri .
8bd0068d 522 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
2b665f28 523
2e25760a 524 if ($type0 =='message' && $type1 == 'rfc822') {
202bcbcc 525 $default_page = $base_uri . 'src/read_body.php';
2e25760a 526 $rfc822_header = $att->rfc822_header;
098ea084 527 $filename = $rfc822_header->subject;
6cc08d8b 528 if (trim( $filename ) == '') {
529 $filename = 'untitled-[' . $ent . ']' ;
08b7f7cc 530 }
2e25760a 531 $from_o = $rfc822_header->from;
532 if (is_object($from_o)) {
04ea844e 533 $from_name = decodeHeader($from_o->getAddress(false));
7e697748 534 } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
535 // something weird happens when a digest message is opened and you return to the digest
536 // now the from object is part of an array. Probably the parseHeader call overwrites the info
537 // retrieved from the bodystructure in a different way. We need to fix this later.
538 // possible starting point, do not fetch header we already have and inspect how
539 // the rfc822_header object behaves.
540 $from_name = decodeHeader($from_o[0]->getAddress(false));
2e25760a 541 } else {
542 $from_name = _("Unknown sender");
f0c4dc12 543 }
d0187bd6 544 $description = _("From").': '.$from_name;
23bcec6f 545 } else {
202bcbcc 546 $default_page = $base_uri . 'src/download.php';
02474e43 547 $filename = $att->getFilename();
f810c0b2 548 if ($header->description) {
098ea084 549 $description = decodeHeader($header->description);
f810c0b2 550 } else {
3d8371be 551 $description = '';
552 }
2e25760a 553 }
554
555 $display_filename = $filename;
3d8371be 556 if (isset($passed_ent_id)) {
557 $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
558 } else {
559 $passed_ent_id_link = '';
560 }
561 $defaultlink = $default_page . "?startMessage=$startMessage"
8bd0068d 562 . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
563 . '&amp;ent_id='.$ent.$passed_ent_id_link;
2e25760a 564 if ($where && $what) {
8bd0068d 565 $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
2e25760a 566 }
7e2ff844 567 // IE does make use of mime content sniffing. Forcing a download
568 // prohibit execution of XSS inside an application/octet-stream attachment
569 if ($type0 == 'application' && $type1 == 'octet-stream') {
570 $defaultlink .= '&amp;absolute_dl=true';
571 }
2b665f28 572
3d8371be 573 /* This executes the attachment hook with a specific MIME-type.
8bd0068d 574 * If that doesn't have results, it tries if there's a rule
9b94c54d 575 * for a more generic type. Finally, a hook for ALL attachment
576 * types is run as well.
8bd0068d 577 */
8dca4d22 578 // First remember the default link.
579 $defaultlink_orig = $defaultlink;
580
d849b570 581 /* The API for this hook has changed as of 1.5.2 so that all plugin
582 arguments are passed in an array instead of each their own plugin
583 argument, and arguments are passed by reference, so instead of
584 returning any changes, changes should simply be made to the original
585 arguments themselves. */
9c3b2d22 586 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
587 &$defaultlink, &$display_filename, &$where, &$what);
588 do_hook("attachment $type0/$type1", $temp);
8dca4d22 589 if(count($links) <= 1 && $defaultlink == $defaultlink_orig) {
d849b570 590 /* The API for this hook has changed as of 1.5.2 so that all plugin
591 arguments are passed in an array instead of each their own plugin
592 argument, and arguments are passed by reference, so instead of
593 returning any changes, changes should simply be made to the original
594 arguments themselves. */
9c3b2d22 595 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
596 &$defaultlink, &$display_filename, &$where, &$what);
597 do_hook("attachment $type0/*", $temp);
2e25760a 598 }
d849b570 599 /* The API for this hook has changed as of 1.5.2 so that all plugin
600 arguments are passed in an array instead of each their own plugin
601 argument, and arguments are passed by reference, so instead of
602 returning any changes, changes should simply be made to the original
603 arguments themselves. */
9c3b2d22 604 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
605 &$defaultlink, &$display_filename, &$where, &$what);
8dca4d22 606 // Do not let a generic plugin change the default link if a more
607 // specialized one already did it...
608 if ($defaultlink != $defaultlink_orig) {
609 $dummy = '';
610 $temp[5] = &$dummy;
611 }
9c3b2d22 612 do_hook("attachment */*", $temp);
77b88425 613
d67f519a 614 $this_attachment = array();
615 $this_attachment['Name'] = decodeHeader($display_filename);
616 $this_attachment['Description'] = $description;
617 $this_attachment['DefaultHREF'] = $defaultlink;
618 $this_attachment['DownloadHREF'] = $links['download link']['href'];
619 $this_attachment['ViewHREF'] = isset($links['attachment_common']) ? $links['attachment_common']['href'] : '';
620 $this_attachment['Size'] = $header->size;
621 $this_attachment['ContentType'] = htmlspecialchars($type0 .'/'. $type1);
622 $this_attachment['OtherLinks'] = array();
3d8371be 623 foreach ($links as $val) {
d0187bd6 624 if ($val['text']==_("Download") || $val['text'] == _("View"))
625 continue;
626 if (empty($val['text']) && empty($val['extra']))
627 continue;
2b665f28 628
d67f519a 629 $temp = array();
630 $temp['HREF'] = $val['href'];
631 $temp['Text'] = (empty($val['text']) ? '' : $val['text']) . (empty($val['extra']) ? '' : $val['extra']);
632 $this_attachment['OtherLinks'][] = $temp;
2e25760a 633 }
d67f519a 634 $attachments[] = $this_attachment;
2b665f28 635
3d8371be 636 unset($links);
2e25760a 637 }
2b665f28 638
d67f519a 639 return $attachments;
640}
641
642/**
643 * Displays attachment links and information
644 *
645 * Since 1.3.0 function is not included in formatBody() call.
646 *
647 * Since 1.0.2 uses attachment $type0/$type1 hook.
648 * Since 1.2.5 uses attachment $type0/* hook.
649 * Since 1.5.0 uses attachments_bottom hook.
650 * Since 1.5.2 uses templates and does *not* return a value.
651 *
652 * @param object $message SquirrelMail message object
653 * @param array $exclude_id message parts that are not attachments.
654 * @param string $mailbox mailbox name
655 * @param integer $id message id
656 */
657function formatAttachments($message, $exclude_id, $mailbox, $id) {
658 global $oTemplate;
2b665f28 659
d67f519a 660 $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
d0187bd6 661
662 $oTemplate->assign('attachments', $attach);
663 $oTemplate->display('read_attachments.tpl');
451f74a2 664}
b74ba498 665
7c7b74b3 666function sqimap_base64_decode(&$string) {
7c0ec1d8 667
b17a8968 668 // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
7c0ec1d8 669 // fly decoding (to reduce memory usage) you have to check if the
670 // data has incomplete pairs
671
b17a8968 672 // Remove the noise in order to check if the 4 bytes pairs are complete
7c0ec1d8 673 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
674
42ce44f8 675 $sStringRem = '';
7c0ec1d8 676 $iMod = strlen($string) % 4;
677 if ($iMod) {
678 $sStringRem = substr($string,-$iMod);
b17a8968 679 // Check if $sStringRem contains padding characters
7c0ec1d8 680 if (substr($sStringRem,-1) != '=') {
681 $string = substr($string,0,-$iMod);
682 } else {
683 $sStringRem = '';
684 }
685 }
7c7b74b3 686 $string = base64_decode($string);
7c0ec1d8 687 return $sStringRem;
7c7b74b3 688}
689
fdf7cef1 690/**
fa26ab45 691 * Decodes encoded string (usually message body)
692 *
693 * This function decodes a string (usually the message body)
694 * depending on the encoding type. Currently quoted-printable
695 * and base64 encodings are supported.
696 *
697 * The decode_body hook was added to this function in 1.4.2/1.5.0.
698 * The $force_crlf parameter was added in 1.5.2.
699 *
700 * @param string $string The encoded string
701 * @param string $encoding used encoding
702 * @param string $force_crlf Whether or not to force CRLF or LF
703 * line endings (or to leave as is).
704 * If given as "LF", line endings will
705 * all be converted to LF; if "CRLF",
706 * line endings will all be converted
707 * to CRLF. If given as an empty value,
708 * the global $default_force_crlf will
709 * be consulted (it can be specified in
710 * config/config_local.php). Otherwise,
711 * any other value will cause the string
712 * to be left alone. Note that this will
713 * be overridden to "LF" if not using at
714 * least PHP version 4.3.0. (OPTIONAL;
715 * default is empty - consult global
716 * default value)
717 *
718 * @return string The decoded string
fdf7cef1 719 *
fdf7cef1 720 * @since 1.0
fa26ab45 721 *
fdf7cef1 722 */
fa26ab45 723function decodeBody($string, $encoding, $force_crlf='') {
724
725 global $force_crlf_default;
726 if (empty($force_crlf)) $force_crlf = $force_crlf_default;
727 $force_crlf = strtoupper($force_crlf);
728
729 // must force line endings to LF due to broken
730 // quoted_printable_decode() in PHP versions
731 // before 4.3.0 (see below)
732 //
733 if (!check_php_version(4, 3, 0) || $force_crlf == 'LF')
734 $string = str_replace("\r\n", "\n", $string);
735 else if ($force_crlf == 'CRLF')
736 $string = str_replace("\n", "\r\n", $string);
83be314a 737
b583c3e8 738 $encoding = strtolower($encoding);
3d8371be 739
d849b570 740 $encoding_handler = do_hook('decode_body', $encoding);
5166f86a 741
742
fa26ab45 743 // plugins get first shot at decoding the string
5166f86a 744 //
745 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
fa26ab45 746 $string = $encoding_handler('decode', $string);
5166f86a 747
fdf7cef1 748 } elseif ($encoding == 'quoted-printable' ||
8bd0068d 749 $encoding == 'quoted_printable') {
fa26ab45 750
751 // quoted_printable_decode() function is broken in older
752 // php versions. Text with \r\n decoding was fixed only
753 // in php 4.3.0. Minimal code requirement is PHP 4.0.4+
754 // and the above call to: str_replace("\r\n", "\n", $string);
755 //
756 $string = quoted_printable_decode($string);
757
fdf7cef1 758 } elseif ($encoding == 'base64') {
fa26ab45 759 $string = base64_decode($string);
b583c3e8 760 }
3d8371be 761
b583c3e8 762 // All other encodings are returned raw.
fa26ab45 763 return $string;
451f74a2 764}
765
9f7f68c3 766/**
8bd0068d 767 * Decodes headers
768 *
e1115979 769 * This function decodes strings that are encoded according to
8bd0068d 770 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
771 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
772 *
773 * @param string $string header string that has to be made readable
774 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
bcae324b 775 * @param boolean $htmlsafe preserve spaces and sanitize html special characters. defaults to true
8bd0068d 776 * @param boolean $decide decide if string can be utfencoded. defaults to false
777 * @return string decoded header string
778 */
bcae324b 779function decodeHeader ($string, $utfencode=true,$htmlsafe=true,$decide=false) {
d6f584fc 780 global $languages, $squirrelmail_language,$default_charset;
79e07c7e 781 if (is_array($string)) {
782 $string = implode("\n", $string);
783 }
da2415c1 784
10dec454 785 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 786 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
33a55f5a 787 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
08b7f7cc 788 // Do we need to return at this point?
789 // return $string;
83be314a 790 }
79e07c7e 791 $i = 0;
08b7f7cc 792 $iLastMatch = -2;
db65b6b0 793 $encoded = true;
0a06275a 794
098ea084 795 $aString = explode(' ',$string);
08b7f7cc 796 $ret = '';
098ea084 797 foreach ($aString as $chunk) {
358a78a1 798 if ($encoded && $chunk === '') {
08b7f7cc 799 continue;
358a78a1 800 } elseif ($chunk === '') {
08b7f7cc 801 $ret .= ' ';
802 continue;
803 }
098ea084 804 $encoded = false;
08b7f7cc 805 /* if encoded words are not separated by a linear-space-white we still catch them */
806 $j = $i-1;
7e6ca3e8 807
08b7f7cc 808 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
809 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
810 if ($iLastMatch !== $j) {
bcae324b 811 if ($htmlsafe) {
9f7f68c3 812 $ret .= '&#32;';
08b7f7cc 813 } else {
814 $ret .= ' ';
815 }
816 }
817 $iLastMatch = $i;
818 $j = $i;
bcae324b 819 if ($htmlsafe) {
cb718de0 820 $ret .= htmlspecialchars($res[1]);
821 } else {
822 $ret .= $res[1];
823 }
098ea084 824 $encoding = ucfirst($res[3]);
d6f584fc 825
826 /* decide about valid decoding */
827 if ($decide && is_conversion_safe($res[2])) {
8bd0068d 828 $utfencode=true;
829 $can_be_encoded=true;
d6f584fc 830 } else {
8bd0068d 831 $can_be_encoded=false;
d6f584fc 832 }
098ea084 833 switch ($encoding)
834 {
8bd0068d 835 case 'B':
836 $replace = base64_decode($res[4]);
837 if ($utfencode) {
838 if ($can_be_encoded) {
839 /* convert string to different charset,
840 * if functions asks for it (usually in compose)
841 */
bcae324b 842 $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsafe);
8bd0068d 843 } else {
844 // convert string to html codes in order to display it
845 $ret .= charset_decode($res[2],$replace);
846 }
fab65ca9 847 } else {
bcae324b 848 if ($htmlsafe) {
8bd0068d 849 $replace = htmlspecialchars($replace);
850 }
851 $ret.= $replace;
fab65ca9 852 }
8bd0068d 853 break;
854 case 'Q':
855 $replace = str_replace('_', ' ', $res[4]);
856 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
857 $replace);
858 if ($utfencode) {
859 if ($can_be_encoded) {
860 /* convert string to different charset,
861 * if functions asks for it (usually in compose)
862 */
bcae324b 863 $replace = charset_convert($res[2], $replace,$default_charset,$htmlsafe);
8bd0068d 864 } else {
865 // convert string to html codes in order to display it
866 $replace = charset_decode($res[2], $replace);
867 }
868 } else {
bcae324b 869 if ($htmlsafe) {
8bd0068d 870 $replace = htmlspecialchars($replace);
871 }
098ea084 872 }
8bd0068d 873 $ret .= $replace;
874 break;
875 default:
876 break;
79e07c7e 877 }
098ea084 878 $chunk = $res[5];
879 $encoded = true;
08b7f7cc 880 }
881 if (!$encoded) {
bcae324b 882 if ($htmlsafe) {
9f7f68c3 883 $ret .= '&#32;';
08b7f7cc 884 } else {
885 $ret .= ' ';
da2415c1 886 }
08b7f7cc 887 }
dc3d13a7 888
bcae324b 889 if (!$encoded && $htmlsafe) {
dc3d13a7 890 $ret .= htmlspecialchars($chunk);
891 } else {
892 $ret .= $chunk;
893 }
098ea084 894 ++$i;
895 }
fd81e884 896 /* remove the first added space */
897 if ($ret) {
bcae324b 898 if ($htmlsafe) {
9f7f68c3 899 $ret = substr($ret,5);
fd81e884 900 } else {
901 $ret = substr($ret,1);
902 }
903 }
da2415c1 904
08b7f7cc 905 return $ret;
451f74a2 906}
907
9f7f68c3 908/**
a24cf710 909 * Encodes header
8bd0068d 910 *
a24cf710 911 * Function uses XTRA_CODE _encodeheader function, if such function exists.
a24cf710 912 *
758a7889 913 * Function uses Q encoding by default and encodes a string according to RFC
914 * 1522 for use in headers if it contains 8-bit characters or anything that
a24cf710 915 * looks like it should be encoded.
8bd0068d 916 *
758a7889 917 * Function switches to B encoding and encodeHeaderBase64() function, if
918 * string is 8bit and multibyte character set supported by mbstring extension
919 * is used. It can cause E_USER_NOTICE errors, if interface is used with
f270a6eb 920 * multibyte character set unsupported by mbstring extension.
921 *
8bd0068d 922 * @param string $string header string, that has to be encoded
923 * @return string quoted-printable encoded string
f270a6eb 924 * @todo make $mb_charsets system wide constant
8bd0068d 925 */
451f74a2 926function encodeHeader ($string) {
6fbd125b 927 global $default_charset, $languages, $squirrelmail_language;
83be314a 928
10dec454 929 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 930 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
33a55f5a 931 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
83be314a 932 }
793cc001 933
a24cf710 934 // Use B encoding for multibyte charsets
f270a6eb 935 $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
936 if (in_array($default_charset,$mb_charsets) &&
937 in_array($default_charset,sq_mb_list_encodings()) &&
938 sq_is8bit($string)) {
939 return encodeHeaderBase64($string,$default_charset);
940 } elseif (in_array($default_charset,$mb_charsets) &&
941 sq_is8bit($string) &&
942 ! in_array($default_charset,sq_mb_list_encodings())) {
943 // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
944 // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
945 }
a24cf710 946
451f74a2 947 // Encode only if the string contains 8-bit characters or =?
3d8371be 948 $j = strlen($string);
098ea084 949 $max_l = 75 - strlen($default_charset) - 7;
950 $aRet = array();
451f74a2 951 $ret = '';
c96c32f4 952 $iEncStart = $enc_init = false;
0d53f0f9 953 $cur_l = $iOffset = 0;
3d8371be 954 for($i = 0; $i < $j; ++$i) {
c96c32f4 955 switch($string{$i})
956 {
8bd0068d 957 case '=':
958 case '<':
959 case '>':
960 case ',':
961 case '?':
962 case '_':
963 if ($iEncStart === false) {
964 $iEncStart = $i;
965 }
966 $cur_l+=3;
967 if ($cur_l > ($max_l-2)) {
968 /* if there is an stringpart that doesn't need encoding, add it */
08b7f7cc 969 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
c96c32f4 970 $aRet[] = "=?$default_charset?Q?$ret?=";
971 $iOffset = $i;
972 $cur_l = 0;
973 $ret = '';
974 $iEncStart = false;
08b7f7cc 975 } else {
8bd0068d 976 $ret .= sprintf("=%02X",ord($string{$i}));
c96c32f4 977 }
8bd0068d 978 break;
979 case '(':
980 case ')':
981 if ($iEncStart !== false) {
08b7f7cc 982 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
8bd0068d 983 $aRet[] = "=?$default_charset?Q?$ret?=";
c96c32f4 984 $iOffset = $i;
8bd0068d 985 $cur_l = 0;
986 $ret = '';
987 $iEncStart = false;
c96c32f4 988 }
8bd0068d 989 break;
990 case ' ':
c96c32f4 991 if ($iEncStart !== false) {
098ea084 992 $cur_l++;
993 if ($cur_l > $max_l) {
08b7f7cc 994 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
c96c32f4 995 $aRet[] = "=?$default_charset?Q?$ret?=";
c96c32f4 996 $iOffset = $i;
997 $cur_l = 0;
098ea084 998 $ret = '';
8bd0068d 999 $iEncStart = false;
08b7f7cc 1000 } else {
8bd0068d 1001 $ret .= '_';
c96c32f4 1002 }
3d8371be 1003 }
8bd0068d 1004 break;
1005 default:
1006 $k = ord($string{$i});
1007 if ($k > 126) {
1008 if ($iEncStart === false) {
1009 // do not start encoding in the middle of a string, also take the rest of the word.
1010 $sLeadString = substr($string,0,$i);
1011 $aLeadString = explode(' ',$sLeadString);
1012 $sToBeEncoded = array_pop($aLeadString);
1013 $iEncStart = $i - strlen($sToBeEncoded);
1014 $ret .= $sToBeEncoded;
1015 $cur_l += strlen($sToBeEncoded);
1016 }
1017 $cur_l += 3;
1018 /* first we add the encoded string that reached it's max size */
1019 if ($cur_l > ($max_l-2)) {
1020 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1021 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
1022 $cur_l = 3;
1023 $ret = '';
1024 $iOffset = $i;
1025 $iEncStart = $i;
1026 }
1027 $enc_init = true;
1028 $ret .= sprintf("=%02X", $k);
1029 } else {
1030 if ($iEncStart !== false) {
1031 $cur_l++;
1032 if ($cur_l > $max_l) {
1033 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1034 $aRet[] = "=?$default_charset?Q?$ret?=";
1035 $iEncStart = false;
1036 $iOffset = $i;
1037 $cur_l = 0;
1038 $ret = '';
1039 } else {
1040 $ret .= $string{$i};
1041 }
1042 }
1043 }
1044 break;
f7b3ba37 1045 }
451f74a2 1046 }
793cc001 1047
c96c32f4 1048 if ($enc_init) {
1049 if ($iEncStart !== false) {
1050 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1051 $aRet[] = "=?$default_charset?Q?$ret?=";
1052 } else {
1053 $aRet[] = substr($string,$iOffset);
1054 }
1055 $string = implode('',$aRet);
451f74a2 1056 }
3d8371be 1057 return $string;
451f74a2 1058}
b74ba498 1059
f270a6eb 1060/**
1061 * Encodes string according to rfc2047 B encoding header formating rules
1062 *
758a7889 1063 * It is recommended way to encode headers with character sets that store
f270a6eb 1064 * symbols in more than one byte.
1065 *
1066 * Function requires mbstring support. If required mbstring functions are missing,
1067 * function returns false and sets E_USER_WARNING level error message.
1068 *
758a7889 1069 * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
1070 * that mbstring functions will generate E_WARNING errors, if unsupported
f270a6eb 1071 * character set is used. mb_encode_mimeheader function provided by php
1072 * mbstring extension is not used in order to get better control of header
1073 * encoding.
1074 *
758a7889 1075 * Used php code functions - function_exists(), trigger_error(), strlen()
1076 * (is used with charset names and base64 strings). Used php mbstring
f270a6eb 1077 * functions - mb_strlen and mb_substr.
1078 *
758a7889 1079 * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
f270a6eb 1080 * encoding), rfc 2822 (header folding)
1081 *
1082 * @param string $string header string that must be encoded
758a7889 1083 * @param string $charset character set. Must be supported by mbstring extension.
f270a6eb 1084 * Use sq_mb_list_encodings() to detect supported charsets.
1085 * @return string string encoded according to rfc2047 B encoding formating rules
1086 * @since 1.5.1
1087 * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1088 * @todo Do we want to control max length of header?
1089 * @todo Do we want to control EOL (end-of-line) marker?
1090 * @todo Do we want to translate error message?
1091 */
1092function encodeHeaderBase64($string,$charset) {
1093 /**
1094 * Check mbstring function requirements.
1095 */
1096 if (! function_exists('mb_strlen') ||
1097 ! function_exists('mb_substr')) {
1098 // set E_USER_WARNING
1099 trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
1100 // return false
1101 return false;
1102 }
1103
1104 // initial return array
1105 $aRet = array();
1106
1107 /**
1108 * header length = 75 symbols max (same as in encodeHeader)
1109 * remove $charset length
1110 * remove =? ? ?= (5 chars)
1111 * remove 2 more chars (\r\n ?)
1112 */
1113 $iMaxLength = 75 - strlen($charset) - 7;
1114
1115 // set first character position
1116 $iStartCharNum = 0;
1117
1118 // loop through all characters. count characters and not bytes.
1119 for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1120 // encode string from starting character to current character.
1121 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1122
1123 // Check encoded string length
1124 if(strlen($encoded_string)>$iMaxLength) {
1125 // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1126 $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1127
1128 // set new starting character
1129 $iStartCharNum = $iCharNum-1;
1130
1131 // encode last char (in case it is last character in string)
1132 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1133 } // if string is shorter than max length - add next character
1134 }
1135
1136 // add last encoded string to array
1137 $aRet[] = $encoded_string;
1138
1139 // set initial return string
1140 $sRet = '';
1141
1142 // loop through encoded strings
1143 foreach($aRet as $string) {
1144 // TODO: Do we want to control EOL (end-of-line) marker
1145 if ($sRet!='') $sRet.= " ";
1146
1147 // add header tags and encoded string to return string
1148 $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1149 }
1150
1151 return $sRet;
1152}
1153
691a2d25 1154/* This function trys to locate the entity_id of a specific mime element */
3d8371be 1155function find_ent_id($id, $message) {
a171b359 1156 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1157 if ($message->entities[$i]->header->type0 == 'multipart') {
3d8371be 1158 $ret = find_ent_id($id, $message->entities[$i]);
451f74a2 1159 } else {
3d8371be 1160 if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
d8cffbab 1161// if (sq_check_save_extension($message->entities[$i])) {
8bd0068d 1162 return $message->entities[$i]->entity_id;
da2415c1 1163// }
c8f5f606 1164 } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1165 /**
1166 * This is part of a fix for Outlook Express 6.x generating
1167 * cid URLs without creating content-id headers
1168 * @@JA - 20050207
1169 */
1170 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1171 return $message->entities[$i]->entity_id;
1172 }
3d8371be 1173 }
a3daaaf3 1174 }
451f74a2 1175 }
3d8371be 1176 return $ret;
451f74a2 1177}
a3daaaf3 1178
e5e9381a 1179function sq_check_save_extension($message) {
1180 $filename = $message->getFilename();
1181 $ext = substr($filename, strrpos($filename,'.')+1);
1182 $save_extensions = array('jpg','jpeg','gif','png','bmp');
3d8371be 1183 return in_array($ext, $save_extensions);
e5e9381a 1184}
1185
1186
691a2d25 1187/**
8bd0068d 1188 ** HTMLFILTER ROUTINES
1189 */
451f74a2 1190
2dd879b8 1191/**
0493ed11 1192 * This function checks attribute values for entity-encoded values
1193 * and returns them translated into 8-bit strings so we can run
1194 * checks on them.
8bd0068d 1195 *
0493ed11 1196 * @param $attvalue A string to run entity check against.
1197 * @return Nothing, modifies a reference value.
8bd0068d 1198 */
0493ed11 1199function sq_defang(&$attvalue){
1200 $me = 'sq_defang';
2dd879b8 1201 /**
0493ed11 1202 * Skip this if there aren't ampersands or backslashes.
8bd0068d 1203 */
0493ed11 1204 if (strpos($attvalue, '&') === false
1205 && strpos($attvalue, '\\') === false){
1206 return;
2dd879b8 1207 }
0493ed11 1208 $m = false;
2b665f28 1209 // before deent, translate the dangerous unicode characters and ... to safe values
1210 // otherwise the regular expressions do not match.
1211
1212
1213
0493ed11 1214 do {
1215 $m = false;
1216 $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1217 $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1218 $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1219 } while ($m == true);
1220 $attvalue = stripslashes($attvalue);
2dd879b8 1221}
1222
1223/**
8bd0068d 1224 * Kill any tabs, newlines, or carriage returns. Our friends the
1225 * makers of the browser with 95% market value decided that it'd
1226 * be funny to make "java[tab]script" be just as good as "javascript".
1227 *
1228 * @param attvalue The attribute value before extraneous spaces removed.
0493ed11 1229 * @return attvalue Nothing, modifies a reference value.
8bd0068d 1230 */
0493ed11 1231function sq_unspace(&$attvalue){
1232 $me = 'sq_unspace';
1233 if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1234 $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1235 Array('', '', '', '', ''), $attvalue);
2dd879b8 1236 }
2dd879b8 1237}
1238
2b665f28 1239/**
88de4926 1240 * Translate all dangerous Unicode or Shift_JIS characters which are accepted by
2b665f28 1241 * IE as regular characters.
1242 *
1243 * @param attvalue The attribute value before dangerous characters are translated.
1244 * @return attvalue Nothing, modifies a reference value.
1245 * @author Marc Groot Koerkamp.
1246 */
1247function sq_fixIE_idiocy(&$attvalue) {
1248 // remove NUL
1249 $attvalue = str_replace("\0", "", $attvalue);
1250 // remove comments
1251 $attvalue = preg_replace("/(\/\*.*?\*\/)/","",$attvalue);
1252
88de4926 1253 // IE has the evil habit of accepting every possible value for the attribute expression.
1254 // The table below contains characters which are parsed by IE if they are used in the "expression"
2b665f28 1255 // attribute value.
1256 $aDangerousCharsReplacementTable = array(
1257 array('&#x029F;', '&#0671;' ,/* L UNICODE IPA Extension */
1258 '&#x0280;', '&#0640;' ,/* R UNICODE IPA Extension */
1259 '&#x0274;', '&#0628;' ,/* N UNICODE IPA Extension */
567dc524 1260 '&#xFF25;', '&#65317;' ,/* Unicode FULLWIDTH LATIN CAPITAL LETTER E */
1261 '&#xFF45;', '&#65349;' ,/* Unicode FULLWIDTH LATIN SMALL LETTER E */
2b665f28 1262 '&#xFF38;', '&#65336;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER X */
1263 '&#xFF58;', '&#65368;',/* Unicode FULLWIDTH LATIN SMALL LETTER X */
1264 '&#xFF30;', '&#65328;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER P */
1265 '&#xFF50;', '&#65360;',/* Unicode FULLWIDTH LATIN SMALL LETTER P */
1266 '&#xFF32;', '&#65330;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER R */
1267 '&#xFF52;', '&#65362;',/* Unicode FULLWIDTH LATIN SMALL LETTER R */
1268 '&#xFF33;', '&#65331;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER S */
1269 '&#xFF53;', '&#65363;',/* Unicode FULLWIDTH LATIN SMALL LETTER S */
1270 '&#xFF29;', '&#65321;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER I */
1271 '&#xFF49;', '&#65353;',/* Unicode FULLWIDTH LATIN SMALL LETTER I */
1272 '&#xFF2F;', '&#65327;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER O */
1273 '&#xFF4F;', '&#65359;',/* Unicode FULLWIDTH LATIN SMALL LETTER O */
1274 '&#xFF2E;', '&#65326;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER N */
1275 '&#xFF4E;', '&#65358;',/* Unicode FULLWIDTH LATIN SMALL LETTER N */
1276 '&#xFF2C;', '&#65324;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER L */
1277 '&#xFF4C;', '&#65356;',/* Unicode FULLWIDTH LATIN SMALL LETTER L */
1278 '&#xFF35;', '&#65333;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER U */
1279 '&#xFF55;', '&#65365;',/* Unicode FULLWIDTH LATIN SMALL LETTER U */
1280 '&#x207F;', '&#8319;' ,/* Unicode SUPERSCRIPT LATIN SMALL LETTER N */
567dc524 1281 "\xEF\xBC\xA5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER E */ // in unicode this is some Chinese char range
1282 "\xEF\xBD\x85", /* Shift JIS FULLWIDTH LATIN SMALL LETTER E */
1283 "\xEF\xBC\xB8", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER X */
1284 "\xEF\xBD\x98", /* Shift JIS FULLWIDTH LATIN SMALL LETTER X */
1285 "\xEF\xBC\xB0", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER P */
1286 "\xEF\xBD\x90", /* Shift JIS FULLWIDTH LATIN SMALL LETTER P */
1287 "\xEF\xBC\xB2", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER R */
1288 "\xEF\xBD\x92", /* Shift JIS FULLWIDTH LATIN SMALL LETTER R */
1289 "\xEF\xBC\xB3", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER S */
1290 "\xEF\xBD\x93", /* Shift JIS FULLWIDTH LATIN SMALL LETTER S */
1291 "\xEF\xBC\xA9", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER I */
1292 "\xEF\xBD\x89", /* Shift JIS FULLWIDTH LATIN SMALL LETTER I */
1293 "\xEF\xBC\xAF", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER O */
1294 "\xEF\xBD\x8F", /* Shift JIS FULLWIDTH LATIN SMALL LETTER O */
1295 "\xEF\xBC\xAE", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER N */
1296 "\xEF\xBD\x8E", /* Shift JIS FULLWIDTH LATIN SMALL LETTER N */
1297 "\xEF\xBC\xAC", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER L */
1298 "\xEF\xBD\x8C", /* Shift JIS FULLWIDTH LATIN SMALL LETTER L */
1299 "\xEF\xBC\xB5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER U */
1300 "\xEF\xBD\x95", /* Shift JIS FULLWIDTH LATIN SMALL LETTER U */
1301 "\xE2\x81\xBF", /* Shift JIS FULLWIDTH SUPERSCRIPT N */
1302 "\xCA\x9F", /* L UNICODE IPA Extension */
1303 "\xCA\x80", /* R UNICODE IPA Extension */
1304 "\xC9\xB4"), /* N UNICODE IPA Extension */
2b665f28 1305 array('l', 'l', 'r','r','n','n',
567dc524 1306 'E','E','e','e','X','X','x','x','P','P','p','p','R','R','r','r','S','S','s','s','I','I',
1307 'i','i','O','O','o','o','N','N','n','n','L','L','l','l','U','U','u','u','n','n',
1308 '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 1309 $attvalue = str_replace($aDangerousCharsReplacementTable[0],$aDangerousCharsReplacementTable[1],$attvalue);
1310
88de4926 1311 // Escapes are useful for special characters like "{}[]()'&. In other cases they are
1312 // used for XSS.
2b665f28 1313 $attvalue = preg_replace("/(\\\\)([a-zA-Z]{1})/",'$2',$attvalue);
1314}
1315
691a2d25 1316/**
8bd0068d 1317 * This function returns the final tag out of the tag name, an array
1318 * of attributes, and the type of the tag. This function is called by
1319 * sq_sanitize internally.
1320 *
1321 * @param $tagname the name of the tag.
1322 * @param $attary the array of attributes and their values
1323 * @param $tagtype The type of the tag (see in comments).
1324 * @return a string with the final tag representation.
1325 */
691a2d25 1326function sq_tagprint($tagname, $attary, $tagtype){
b583c3e8 1327 $me = 'sq_tagprint';
3d8371be 1328
691a2d25 1329 if ($tagtype == 2){
1330 $fulltag = '</' . $tagname . '>';
1331 } else {
1332 $fulltag = '<' . $tagname;
1333 if (is_array($attary) && sizeof($attary)){
1334 $atts = Array();
1335 while (list($attname, $attvalue) = each($attary)){
1336 array_push($atts, "$attname=$attvalue");
1337 }
1338 $fulltag .= ' ' . join(" ", $atts);
1339 }
1340 if ($tagtype == 3){
b583c3e8 1341 $fulltag .= ' /';
691a2d25 1342 }
b583c3e8 1343 $fulltag .= '>';
451f74a2 1344 }
691a2d25 1345 return $fulltag;
451f74a2 1346}
a3daaaf3 1347
691a2d25 1348/**
8bd0068d 1349 * A small helper function to use with array_walk. Modifies a by-ref
1350 * value and makes it lowercase.
1351 *
1352 * @param $val a value passed by-ref.
1353 * @return void since it modifies a by-ref value.
1354 */
691a2d25 1355function sq_casenormalize(&$val){
1356 $val = strtolower($val);
1357}
451f74a2 1358
691a2d25 1359/**
8bd0068d 1360 * This function skips any whitespace from the current position within
1361 * a string and to the next non-whitespace value.
1362 *
1363 * @param $body the string
1364 * @param $offset the offset within the string where we should start
1365 * looking for the next non-whitespace character.
1366 * @return the location within the $body where the next
1367 * non-whitespace char is located.
1368 */
691a2d25 1369function sq_skipspace($body, $offset){
b583c3e8 1370 $me = 'sq_skipspace';
3d8371be 1371 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
691a2d25 1372 if (sizeof($matches{1})){
1373 $count = strlen($matches{1});
1374 $offset += $count;
451f74a2 1375 }
691a2d25 1376 return $offset;
451f74a2 1377}
a3daaaf3 1378
691a2d25 1379/**
8bd0068d 1380 * This function looks for the next character within a string. It's
1381 * really just a glorified "strpos", except it catches if failures
1382 * nicely.
1383 *
1384 * @param $body The string to look for needle in.
1385 * @param $offset Start looking from this position.
1386 * @param $needle The character/string to look for.
1387 * @return location of the next occurance of the needle, or
1388 * strlen($body) if needle wasn't found.
1389 */
691a2d25 1390function sq_findnxstr($body, $offset, $needle){
3d8371be 1391 $me = 'sq_findnxstr';
691a2d25 1392 $pos = strpos($body, $needle, $offset);
1393 if ($pos === FALSE){
1394 $pos = strlen($body);
451f74a2 1395 }
691a2d25 1396 return $pos;
451f74a2 1397}
a3daaaf3 1398
691a2d25 1399/**
8bd0068d 1400 * This function takes a PCRE-style regexp and tries to match it
1401 * within the string.
1402 *
1403 * @param $body The string to look for needle in.
1404 * @param $offset Start looking from here.
1405 * @param $reg A PCRE-style regex to match.
1406 * @return Returns a false if no matches found, or an array
1407 * with the following members:
1408 * - integer with the location of the match within $body
1409 * - string with whatever content between offset and the match
1410 * - string with whatever it is we matched
1411 */
691a2d25 1412function sq_findnxreg($body, $offset, $reg){
b583c3e8 1413 $me = 'sq_findnxreg';
691a2d25 1414 $matches = Array();
1415 $retarr = Array();
7d06541f 1416 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1417 if (!isset($matches{0}) || !$matches{0}){
691a2d25 1418 $retarr = false;
1419 } else {
1420 $retarr{0} = $offset + strlen($matches{1});
1421 $retarr{1} = $matches{1};
1422 $retarr{2} = $matches{2};
1423 }
1424 return $retarr;
1425}
a3daaaf3 1426
691a2d25 1427/**
8bd0068d 1428 * This function looks for the next tag.
1429 *
1430 * @param $body String where to look for the next tag.
1431 * @param $offset Start looking from here.
1432 * @return false if no more tags exist in the body, or
1433 * an array with the following members:
1434 * - string with the name of the tag
1435 * - array with attributes and their values
1436 * - integer with tag type (1, 2, or 3)
1437 * - integer where the tag starts (starting "<")
1438 * - integer where the tag ends (ending ">")
1439 * first three members will be false, if the tag is invalid.
1440 */
691a2d25 1441function sq_getnxtag($body, $offset){
b583c3e8 1442 $me = 'sq_getnxtag';
691a2d25 1443 if ($offset > strlen($body)){
1444 return false;
1445 }
1446 $lt = sq_findnxstr($body, $offset, "<");
1447 if ($lt == strlen($body)){
1448 return false;
1449 }
1450 /**
8bd0068d 1451 * We are here:
1452 * blah blah <tag attribute="value">
1453 * \---------^
1454 */
691a2d25 1455 $pos = sq_skipspace($body, $lt+1);
1456 if ($pos >= strlen($body)){
1457 return Array(false, false, false, $lt, strlen($body));
1458 }
1459 /**
8bd0068d 1460 * There are 3 kinds of tags:
1461 * 1. Opening tag, e.g.:
1462 * <a href="blah">
1463 * 2. Closing tag, e.g.:
1464 * </a>
1465 * 3. XHTML-style content-less tag, e.g.:
1466 * <img src="blah" />
1467 */
691a2d25 1468 $tagtype = false;
1469 switch (substr($body, $pos, 1)){
3d8371be 1470 case '/':
1471 $tagtype = 2;
1472 $pos++;
1473 break;
1474 case '!':
1475 /**
8bd0068d 1476 * A comment or an SGML declaration.
1477 */
3d8371be 1478 if (substr($body, $pos+1, 2) == "--"){
1479 $gt = strpos($body, "-->", $pos);
1480 if ($gt === false){
1481 $gt = strlen($body);
1482 } else {
1483 $gt += 2;
1484 }
1485 return Array(false, false, false, $lt, $gt);
bb8d0799 1486 } else {
3d8371be 1487 $gt = sq_findnxstr($body, $pos, ">");
1488 return Array(false, false, false, $lt, $gt);
1489 }
1490 break;
1491 default:
1492 /**
8bd0068d 1493 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1494 * later.
1495 */
3d8371be 1496 $tagtype = 1;
1497 break;
691a2d25 1498 }
a3daaaf3 1499
0493ed11 1500 $tag_start = $pos;
691a2d25 1501 $tagname = '';
1502 /**
8bd0068d 1503 * Look for next [\W-_], which will indicate the end of the tag name.
1504 */
691a2d25 1505 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1506 if ($regary == false){
1507 return Array(false, false, false, $lt, strlen($body));
1508 }
1509 list($pos, $tagname, $match) = $regary;
1510 $tagname = strtolower($tagname);
1511
1512 /**
8bd0068d 1513 * $match can be either of these:
1514 * '>' indicating the end of the tag entirely.
1515 * '\s' indicating the end of the tag name.
1516 * '/' indicating that this is type-3 xhtml tag.
1517 *
1518 * Whatever else we find there indicates an invalid tag.
1519 */
691a2d25 1520 switch ($match){
3d8371be 1521 case '/':
691a2d25 1522 /**
8bd0068d 1523 * This is an xhtml-style tag with a closing / at the
1524 * end, like so: <img src="blah" />. Check if it's followed
1525 * by the closing bracket. If not, then this tag is invalid
1526 */
3d8371be 1527 if (substr($body, $pos, 2) == "/>"){
1528 $pos++;
1529 $tagtype = 3;
1530 } else {
1531 $gt = sq_findnxstr($body, $pos, ">");
1532 $retary = Array(false, false, false, $lt, $gt);
1533 return $retary;
1534 }
1535 case '>':
1536 return Array($tagname, false, $tagtype, $lt, $pos);
1537 break;
1538 default:
1539 /**
8bd0068d 1540 * Check if it's whitespace
1541 */
3d8371be 1542 if (!preg_match('/\s/', $match)){
1543 /**
8bd0068d 1544 * This is an invalid tag! Look for the next closing ">".
1545 */
7d06541f 1546 $gt = sq_findnxstr($body, $lt, ">");
3d8371be 1547 return Array(false, false, false, $lt, $gt);
1548 }
1549 break;
691a2d25 1550 }
3d8371be 1551
691a2d25 1552 /**
8bd0068d 1553 * At this point we're here:
1554 * <tagname attribute='blah'>
1555 * \-------^
1556 *
1557 * At this point we loop in order to find all attributes.
1558 */
691a2d25 1559 $attname = '';
0493ed11 1560 $atttype = false;
691a2d25 1561 $attary = Array();
1562
1563 while ($pos <= strlen($body)){
1564 $pos = sq_skipspace($body, $pos);
1565 if ($pos == strlen($body)){
1566 /**
8bd0068d 1567 * Non-closed tag.
1568 */
691a2d25 1569 return Array(false, false, false, $lt, $pos);
1570 }
1571 /**
8bd0068d 1572 * See if we arrived at a ">" or "/>", which means that we reached
1573 * the end of the tag.
1574 */
691a2d25 1575 $matches = Array();
164800ad 1576 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
c828931c 1577 /**
8bd0068d 1578 * Yep. So we did.
1579 */
c828931c 1580 $pos += strlen($matches{1});
1581 if ($matches{2} == "/>"){
1582 $tagtype = 3;
1583 $pos++;
1584 }
1585 return Array($tagname, $attary, $tagtype, $lt, $pos);
1586 }
a3daaaf3 1587
cca46357 1588 /**
8bd0068d 1589 * There are several types of attributes, with optional
1590 * [:space:] between members.
1591 * Type 1:
1592 * attrname[:space:]=[:space:]'CDATA'
1593 * Type 2:
1594 * attrname[:space:]=[:space:]"CDATA"
1595 * Type 3:
1596 * attr[:space:]=[:space:]CDATA
1597 * Type 4:
1598 * attrname
1599 *
1600 * We leave types 1 and 2 the same, type 3 we check for
1601 * '"' and convert to "&quot" if needed, then wrap in
1602 * double quotes. Type 4 we convert into:
1603 * attrname="yes".
1604 */
3f7c623f 1605 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
691a2d25 1606 if ($regary == false){
1607 /**
8bd0068d 1608 * Looks like body ended before the end of tag.
1609 */
691a2d25 1610 return Array(false, false, false, $lt, strlen($body));
cca46357 1611 }
691a2d25 1612 list($pos, $attname, $match) = $regary;
1613 $attname = strtolower($attname);
1614 /**
8bd0068d 1615 * We arrived at the end of attribute name. Several things possible
1616 * here:
1617 * '>' means the end of the tag and this is attribute type 4
1618 * '/' if followed by '>' means the same thing as above
1619 * '\s' means a lot of things -- look what it's followed by.
1620 * anything else means the attribute is invalid.
1621 */
691a2d25 1622 switch($match){
3d8371be 1623 case '/':
691a2d25 1624 /**
8bd0068d 1625 * This is an xhtml-style tag with a closing / at the
1626 * end, like so: <img src="blah" />. Check if it's followed
1627 * by the closing bracket. If not, then this tag is invalid
1628 */
3d8371be 1629 if (substr($body, $pos, 2) == "/>"){
691a2d25 1630 $pos++;
3d8371be 1631 $tagtype = 3;
691a2d25 1632 } else {
3d8371be 1633 $gt = sq_findnxstr($body, $pos, ">");
1634 $retary = Array(false, false, false, $lt, $gt);
1635 return $retary;
1636 }
1637 case '>':
1638 $attary{$attname} = '"yes"';
1639 return Array($tagname, $attary, $tagtype, $lt, $pos);
1640 break;
1641 default:
1642 /**
8bd0068d 1643 * Skip whitespace and see what we arrive at.
1644 */
3d8371be 1645 $pos = sq_skipspace($body, $pos);
1646 $char = substr($body, $pos, 1);
1647 /**
8bd0068d 1648 * Two things are valid here:
1649 * '=' means this is attribute type 1 2 or 3.
1650 * \w means this was attribute type 4.
1651 * anything else we ignore and re-loop. End of tag and
1652 * invalid stuff will be caught by our checks at the beginning
1653 * of the loop.
1654 */
3d8371be 1655 if ($char == "="){
1656 $pos++;
1657 $pos = sq_skipspace($body, $pos);
691a2d25 1658 /**
8bd0068d 1659 * Here are 3 possibilities:
1660 * "'" attribute type 1
1661 * '"' attribute type 2
1662 * everything else is the content of tag type 3
1663 */
3d8371be 1664 $quot = substr($body, $pos, 1);
1665 if ($quot == "'"){
1666 $regary = sq_findnxreg($body, $pos+1, "\'");
1667 if ($regary == false){
1668 return Array(false, false, false, $lt, strlen($body));
1669 }
1670 list($pos, $attval, $match) = $regary;
1671 $pos++;
1672 $attary{$attname} = "'" . $attval . "'";
1673 } else if ($quot == '"'){
1674 $regary = sq_findnxreg($body, $pos+1, '\"');
1675 if ($regary == false){
1676 return Array(false, false, false, $lt, strlen($body));
1677 }
1678 list($pos, $attval, $match) = $regary;
1679 $pos++;
1680 $attary{$attname} = '"' . $attval . '"';
1681 } else {
1682 /**
8bd0068d 1683 * These are hateful. Look for \s, or >.
1684 */
3d8371be 1685 $regary = sq_findnxreg($body, $pos, "[\s>]");
1686 if ($regary == false){
1687 return Array(false, false, false, $lt, strlen($body));
1688 }
1689 list($pos, $attval, $match) = $regary;
1690 /**
8bd0068d 1691 * If it's ">" it will be caught at the top.
1692 */
3d8371be 1693 $attval = preg_replace("/\"/s", "&quot;", $attval);
1694 $attary{$attname} = '"' . $attval . '"';
7e235a1a 1695 }
3d8371be 1696 } else if (preg_match("|[\w/>]|", $char)) {
691a2d25 1697 /**
8bd0068d 1698 * That was attribute type 4.
1699 */
3d8371be 1700 $attary{$attname} = '"yes"';
1701 } else {
1702 /**
8bd0068d 1703 * An illegal character. Find next '>' and return.
1704 */
3d8371be 1705 $gt = sq_findnxstr($body, $pos, ">");
1706 return Array(false, false, false, $lt, $gt);
451f74a2 1707 }
3d8371be 1708 break;
691a2d25 1709 }
1710 }
1711 /**
8bd0068d 1712 * The fact that we got here indicates that the tag end was never
1713 * found. Return invalid tag indication so it gets stripped.
1714 */
691a2d25 1715 return Array(false, false, false, $lt, strlen($body));
1716}
1717
1718/**
0493ed11 1719 * Translates entities into literal values so they can be checked.
8bd0068d 1720 *
0493ed11 1721 * @param $attvalue the by-ref value to check.
1722 * @param $regex the regular expression to check against.
1723 * @param $hex whether the entites are hexadecimal.
1724 * @return True or False depending on whether there were matches.
8bd0068d 1725 */
0493ed11 1726function sq_deent(&$attvalue, $regex, $hex=false){
b583c3e8 1727 $me = 'sq_deent';
0493ed11 1728 $ret_match = false;
2b665f28 1729 // remove comments
1730 //$attvalue = preg_replace("/(\/\*.*\*\/)/","",$attvalue);
0493ed11 1731 preg_match_all($regex, $attvalue, $matches);
1732 if (is_array($matches) && sizeof($matches[0]) > 0){
1733 $repl = Array();
1734 for ($i = 0; $i < sizeof($matches[0]); $i++){
1735 $numval = $matches[1][$i];
1736 if ($hex){
1737 $numval = hexdec($numval);
a3daaaf3 1738 }
0493ed11 1739 $repl{$matches[0][$i]} = chr($numval);
691a2d25 1740 }
0493ed11 1741 $attvalue = strtr($attvalue, $repl);
1742 return true;
1743 } else {
1744 return false;
691a2d25 1745 }
691a2d25 1746}
1747
1748/**
8bd0068d 1749 * This function runs various checks against the attributes.
1750 *
1751 * @param $tagname String with the name of the tag.
1752 * @param $attary Array with all tag attributes.
1753 * @param $rm_attnames See description for sq_sanitize
1754 * @param $bad_attvals See description for sq_sanitize
1755 * @param $add_attr_to_tag See description for sq_sanitize
1756 * @param $message message object
1757 * @param $id message id
1758 * @return Array with modified attributes.
1759 */
da2415c1 1760function sq_fixatts($tagname,
1761 $attary,
691a2d25 1762 $rm_attnames,
1763 $bad_attvals,
1764 $add_attr_to_tag,
1765 $message,
b3af12ef 1766 $id,
3d8371be 1767 $mailbox
691a2d25 1768 ){
b583c3e8 1769 $me = 'sq_fixatts';
691a2d25 1770 while (list($attname, $attvalue) = each($attary)){
1771 /**
8bd0068d 1772 * See if this attribute should be removed.
1773 */
691a2d25 1774 foreach ($rm_attnames as $matchtag=>$matchattrs){
1775 if (preg_match($matchtag, $tagname)){
1776 foreach ($matchattrs as $matchattr){
1777 if (preg_match($matchattr, $attname)){
1778 unset($attary{$attname});
1779 continue;
1780 }
451f74a2 1781 }
451f74a2 1782 }
691a2d25 1783 }
2b665f28 1784 /**
1785 * Workaround for IE quirks
1786 */
1787 sq_fixIE_idiocy($attvalue);
1788
691a2d25 1789 /**
8bd0068d 1790 * Remove any backslashes, entities, and extraneous whitespace.
1791 */
2b665f28 1792
1793 $oldattvalue = $attvalue;
0493ed11 1794 sq_defang($attvalue);
2b665f28 1795 if ($attname == 'style' && $attvalue !== $oldattvalue) {
1796 // entities are used in the attribute value. In 99% of the cases it's there as XSS
1797 // i.e.<div style="{ left:exp&#x0280;essio&#x0274;( alert('XSS') ) }">
1798 $attvalue = "idiocy";
1799 $attary{$attname} = $attvalue;
1800 }
0493ed11 1801 sq_unspace($attvalue);
af861a34 1802
691a2d25 1803 /**
8bd0068d 1804 * Now let's run checks on the attvalues.
1805 * I don't expect anyone to comprehend this. If you do,
1806 * get in touch with me so I can drive to where you live and
1807 * shake your hand personally. :)
1808 */
691a2d25 1809 foreach ($bad_attvals as $matchtag=>$matchattrs){
1810 if (preg_match($matchtag, $tagname)){
1811 foreach ($matchattrs as $matchattr=>$valary){
1812 if (preg_match($matchattr, $attname)){
1813 /**
8bd0068d 1814 * There are two arrays in valary.
1815 * First is matches.
1816 * Second one is replacements
1817 */
691a2d25 1818 list($valmatch, $valrepl) = $valary;
da2415c1 1819 $newvalue =
691a2d25 1820 preg_replace($valmatch, $valrepl, $attvalue);
1821 if ($newvalue != $attvalue){
1822 $attary{$attname} = $newvalue;
567dc524 1823 $attvalue = $newvalue;
691a2d25 1824 }
1825 }
1826 }
451f74a2 1827 }
a3daaaf3 1828 }
567dc524 1829 if ($attname == 'style') {
1830 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1831 // 8bit and control characters in style attribute values can be used for XSS, remove them
1832 $attary{$attname} = '"disallowed character"';
1833 }
1834 preg_match_all("/url\s*\((.+)\)/si",$attvalue,$aMatch);
1835 if (count($aMatch)) {
1836 foreach($aMatch[1] as $sMatch) {
1837 // url value
1838 $urlvalue = $sMatch;
1839 sq_fix_url($attname, $urlvalue, $message, $id, $mailbox,"'");
1840 $attary{$attname} = str_replace($sMatch,$urlvalue,$attvalue);
1841 }
1842 }
691a2d25 1843 }
ff940ebc 1844 /**
567dc524 1845 * Use white list based filtering on attributes which can contain url's
ff940ebc 1846 */
567dc524 1847 else if ($attname == 'href' || $attname == 'src' || $attname == 'background') {
1848 sq_fix_url($attname, $attvalue, $message, $id, $mailbox);
1849 $attary{$attname} = $attvalue;
ff940ebc 1850 }
a3daaaf3 1851 }
691a2d25 1852 /**
8bd0068d 1853 * See if we need to append any attributes to this tag.
1854 */
691a2d25 1855 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1856 if (preg_match($matchtag, $tagname)){
1857 $attary = array_merge($attary, $addattary);
1858 }
1859 }
1860 return $attary;
451f74a2 1861}
a3daaaf3 1862
567dc524 1863/**
1864 * This function filters url's
1865 *
1866 * @param $attvalue String with attribute value to filter
1867 * @param $message message object
1868 * @param $id message id
1869 * @param $mailbox mailbox
1870 * @param $sQuote quoting characters around url's
1871 */
1872function sq_fix_url($attname, &$attvalue, $message, $id, $mailbox,$sQuote = '"') {
1873 $attvalue = trim($attvalue);
1874 if ($attvalue && ($attvalue[0] =='"'|| $attvalue[0] == "'")) {
1875 // remove the double quotes
1876 $sQuote = $attvalue[0];
1877 $attvalue = trim(substr($attvalue,1,-1));
1878 }
1879
d31f73f1 1880 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
1881 // images off by default.
afbf184c 1882 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET, FALSE);
d31f73f1 1883
567dc524 1884 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1885
1886 /**
1887 * Replace empty src tags with the blank image. src is only used
1888 * for frames, images, and image inputs. Doing a replace should
1889 * not affect them working as should be, however it will stop
1890 * IE from being kicked off when src for img tags are not set
1891 */
1892 if ($attvalue == '') {
1893 $attvalue = '"' . SM_PATH . 'images/blank.png"';
1894 } else {
1895 // first, disallow 8 bit characters and control characters
1896 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1897 switch ($attname) {
1898 case 'href':
1899 $attvalue = $sQuote . 'http://invalid-stuff-detected.example.com' . $sQuote;
1900 break;
1901 default:
1902 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1903 break;
1904 }
1905 } else {
1906 $aUrl = parse_url($attvalue);
1907 if (isset($aUrl['scheme'])) {
1908 switch(strtolower($aUrl['scheme'])) {
d75e755b 1909 case 'mailto':
567dc524 1910 case 'http':
1911 case 'https':
1912 case 'ftp':
1913 if ($attname != 'href') {
1914 if ($view_unsafe_images == false) {
1915 $attvalue = $sQuote . $secremoveimg . $sQuote;
1916 } else {
1917 if (isset($aUrl['path'])) {
8bbbc757 1918
1919 // No one has been able to show that image URIs
1920 // can be exploited, so for now, no restrictions
1921 // are made at all. If this proves to be a problem,
1922 // the commented-out code below can be of help.
1923 // (One consideration is that I see nothing in this
1924 // function that specifically says that we will
1925 // only ever arrive here when inspecting an image
1926 // tag, although that does seem to be the end
1927 // result - e.g., <script src="..."> where malicious
1928 // image URIs are in fact a problem are already
1929 // filtered out elsewhere.
1930 /* ---------------------------------
567dc524 1931 // validate image extension.
1932 $ext = strtolower(substr($aUrl['path'],strrpos($aUrl['path'],'.')));
1933 if (!in_array($ext,array('.jpeg','.jpg','xjpeg','.gif','.bmp','.jpe','.png','.xbm'))) {
8bbbc757 1934 // If URI is to something other than
1935 // a regular image file, get the contents
1936 // and try to see if it is an image.
1937 // Don't use Fileinfo (finfo_file()) because
1938 // we'd need to make the admin configure the
1939 // location of the magic.mime file (FIXME: add finfo_file() support later?)
1940 //
1941 $mime_type = '';
1942 if (function_exists('mime_content_type')
1943 && ($FILE = @fopen($attvalue, 'rb', FALSE))) {
1944
1945 // fetch file
1946 //
1947 $file_contents = '';
1948 while (!feof($FILE)) {
1949 $file_contents .= fread($FILE, 8192);
1950 }
1951 fclose($FILE);
1952
1953 // store file locally
1954 //
1955 global $attachment_dir, $username;
1956 $hashed_attachment_dir = getHashedDir($username, $attachment_dir);
1957 $localfilename = GenerateRandomString(32, '', 7);
1958 $full_localfilename = "$hashed_attachment_dir/$localfilename";
1959 while (file_exists($full_localfilename)) {
1960 $localfilename = GenerateRandomString(32, '', 7);
1961 $full_localfilename = "$hashed_attachment_dir/$localfilename";
1962 }
1963 $FILE = fopen("$hashed_attachment_dir/$localfilename", 'wb');
1964 fwrite($FILE, $file_contents);
1965 fclose($FILE);
1966
1967 // get mime type and remove file
1968 //
1969 $mime_type = mime_content_type("$hashed_attachment_dir/$localfilename");
1970 unlink("$hashed_attachment_dir/$localfilename");
1971 }
1972 // debug: echo "$attvalue FILE TYPE IS $mime_type<HR>";
1973 if (substr(strtolower($mime_type), 0, 5) != 'image') {
1974 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1975 }
567dc524 1976 }
8bbbc757 1977 --------------------------------- */
567dc524 1978 } else {
1979 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1980 }
1981 }
cf088f55 1982 } else {
1983 $attvalue = $sQuote . $attvalue . $sQuote;
567dc524 1984 }
1985 break;
1986 case 'outbind':
1987 /**
1988 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1989 * One day MS might actually make it match something useful, for now, falling
1990 * back to using cid2http, so we can grab the blank.png.
1991 */
cf088f55 1992 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
567dc524 1993 break;
1994 case 'cid':
1995 /**
1996 * Turn cid: urls into http-friendly ones.
1997 */
cf088f55 1998 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
567dc524 1999 break;
2000 default:
2001 $attvalue = $sQuote . SM_PATH . 'images/blank.png' . $sQuote;
2002 break;
2003 }
2004 } else {
2005 if (!(isset($aUrl['path']) && $aUrl['path'] == $secremoveimg)) {
2006 // parse_url did not lead to satisfying result
2007 $attvalue = $sQuote . SM_PATH . 'images/blank.png' . $sQuote;
2008 }
2009 }
2010 }
2011 }
2012}
2013
691a2d25 2014/**
8bd0068d 2015 * This function edits the style definition to make them friendly and
2016 * usable in SquirrelMail.
2017 *
2018 * @param $message the message object
2019 * @param $id the message id
2020 * @param $content a string with whatever is between <style> and </style>
2021 * @param $mailbox the message mailbox
2022 * @return a string with edited content.
2023 */
e60a299a 2024function sq_fixstyle($body, $pos, $message, $id, $mailbox){
b583c3e8 2025 $me = 'sq_fixstyle';
7e2ff844 2026 // workaround for </style> in between comments
2027 $iCurrentPos = $pos;
2028 $content = '';
2029 $sToken = '';
2030 $bSucces = false;
2031 $bEndTag = false;
2032 for ($i=$pos,$iCount=strlen($body);$i<$iCount;++$i) {
2033 $char = $body{$i};
2034 switch ($char) {
2035 case '<':
10c0caeb 2036 $sToken = $char;
7e2ff844 2037 break;
2038 case '/':
2039 if ($sToken == '<') {
2040 $sToken .= $char;
2041 $bEndTag = true;
2042 } else {
2043 $content .= $char;
2044 }
2045 break;
2046 case '>':
2047 if ($bEndTag) {
2048 $sToken .= $char;
2049 if (preg_match('/\<\/\s*style\s*\>/i',$sToken,$aMatch)) {
2050 $newpos = $i + 1;
2051 $bSucces = true;
2052 break 2;
2053 } else {
2054 $content .= $sToken;
2055 }
2056 $bEndTag = false;
2057 } else {
2058 $content .= $char;
2059 }
2060 break;
2061 case '!':
2062 if ($sToken == '<') {
2063 // possible comment
2064 if (isset($body{$i+2}) && substr($body,$i,3) == '!--') {
2065 $i = strpos($body,'-->',$i+3);
2b665f28 2066 if ($i === false) { // no end comment
2067 $i = strlen($body);
2068 }
7e2ff844 2069 $sToken = '';
2070 }
2071 } else {
2072 $content .= $char;
2073 }
2074 break;
2075 default:
2076 if ($bEndTag) {
2077 $sToken .= $char;
2078 } else {
2079 $content .= $char;
2080 }
2081 break;
2082 }
2083 }
2084 if ($bSucces == FALSE){
7d06541f 2085 return array(FALSE, strlen($body));
2086 }
7e2ff844 2087
2b665f28 2088
2089
691a2d25 2090 /**
9ebace19 2091 * First look for general BODY style declaration, which would be
2092 * like so:
2093 * body {background: blah-blah}
2094 * and change it to .bodyclass so we can just assign it to a <div>
2095 */
691a2d25 2096 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
3d8371be 2097 $secremoveimg = '../images/' . _("sec_remove_eng.png");
691a2d25 2098 /**
0493ed11 2099 * Fix url('blah') declarations.
2100 */
2101 // $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
2102 // "url(\\1$secremoveimg\\2)", $content);
2b665f28 2103
567dc524 2104 // first check for 8bit sequences and disallowed control characters
2105 if (preg_match('/[\16-\37\200-\377]+/',$content)) {
2106 $content = '<!-- style block removed by html filter due to presence of 8bit characters -->';
2107 return array($content, $newpos);
2108 }
2109
2b665f28 2110 // IE Sucks hard. We have a special function for it.
2111 sq_fixIE_idiocy($content);
2112
2113 // remove @import line
2114 $content = preg_replace("/^\s*(@import.*)$/mi","\n<!-- @import rules forbidden -->\n",$content);
2115
5b4884be 2116 // translate ur\l and variations (IE parses that)
2b665f28 2117 // TODO check if the sq_fixIE_idiocy function already handles this.
5b4884be 2118 $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
567dc524 2119 preg_match_all("/url\s*\((.+)\)/si",$content,$aMatch);
2120 if (count($aMatch)) {
2121 $aValue = $aReplace = array();
2122 foreach($aMatch[1] as $sMatch) {
2123 // url value
2124 $urlvalue = $sMatch;
2125 sq_fix_url('style',$urlvalue, $message, $id, $mailbox,"'");
2126 $aValue[] = $sMatch;
2127 $aReplace[] = $urlvalue;
0493ed11 2128 }
567dc524 2129 $content = str_replace($aValue,$aReplace,$content);
691a2d25 2130 }
567dc524 2131
9ebace19 2132 /**
2133 * Remove any backslashes, entities, and extraneous whitespace.
2134 */
0493ed11 2135 $contentTemp = $content;
2136 sq_defang($contentTemp);
2137 sq_unspace($contentTemp);
a3daaaf3 2138
691a2d25 2139 /**
8bd0068d 2140 * Fix stupid css declarations which lead to vulnerabilities
2141 * in IE.
2142 */
5db90261 2143 $match = Array('/\/\*.*\*\//',
2144 '/expression/i',
0493ed11 2145 '/behaviou*r/i',
2146 '/binding/i',
2b665f28 2147 '/include-source/i',
2148 '/javascript/i',
2149 '/script/i');
2150 $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy');
0493ed11 2151 $contentNew = preg_replace($match, $replace, $contentTemp);
2152 if ($contentNew !== $contentTemp) {
2153 // insecure css declarations are used. From now on we don't care
2154 // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
2155 $content = $contentNew;
2156 }
7d06541f 2157 return array($content, $newpos);
691a2d25 2158}
a3daaaf3 2159
0493ed11 2160
691a2d25 2161/**
8bd0068d 2162 * This function converts cid: url's into the ones that can be viewed in
2163 * the browser.
2164 *
2165 * @param $message the message object
2166 * @param $id the message id
2167 * @param $cidurl the cid: url.
2168 * @param $mailbox the message mailbox
2169 * @return a string with a http-friendly url
2170 */
b3af12ef 2171function sq_cid2http($message, $id, $cidurl, $mailbox){
691a2d25 2172 /**
8bd0068d 2173 * Get rid of quotes.
2174 */
691a2d25 2175 $quotchar = substr($cidurl, 0, 1);
2dd879b8 2176 if ($quotchar == '"' || $quotchar == "'"){
2177 $cidurl = str_replace($quotchar, "", $cidurl);
2178 } else {
2179 $quotchar = '';
2180 }
691a2d25 2181 $cidurl = substr(trim($cidurl), 4);
0493ed11 2182
2183 $match_str = '/\{.*?\}\//';
2184 $str_rep = '';
2185 $cidurl = preg_replace($match_str, $str_rep, $cidurl);
2186
e5e9381a 2187 $linkurl = find_ent_id($cidurl, $message);
9ebace19 2188 /* in case of non-safe cid links $httpurl should be replaced by a sort of
2189 unsafe link image */
e5e9381a 2190 $httpurl = '';
c8f5f606 2191
8bd0068d 2192 /**
2193 * This is part of a fix for Outlook Express 6.x generating
2194 * cid URLs without creating content-id headers. These images are
2195 * not part of the multipart/related html mail. The html contains
2196 * <img src="cid:{some_id}/image_filename.ext"> references to
2197 * attached images with as goal to render them inline although
2198 * the attachment disposition property is not inline.
2199 */
c8f5f606 2200
2201 if (empty($linkurl)) {
2202 if (preg_match('/{.*}\//', $cidurl)) {
2203 $cidurl = preg_replace('/{.*}\//','', $cidurl);
2204 if (!empty($cidurl)) {
2205 $linkurl = find_ent_id($cidurl, $message);
2206 }
2207 }
2208 }
f8a1ed5a 2209
c8f5f606 2210 if (!empty($linkurl)) {
fa26ab45 2211 $httpurl = $quotchar . sqm_baseuri() . 'src/download.php?absolute_dl=true&amp;' .
8bd0068d 2212 "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
2213 '&amp;ent_id=' . $linkurl . $quotchar;
c8f5f606 2214 } else {
2215 /**
2216 * If we couldn't generate a proper img url, drop in a blank image
2217 * instead of sending back empty, otherwise it causes unusual behaviour
2218 */
bc017c1d 2219 $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
e5e9381a 2220 }
f8a1ed5a 2221
691a2d25 2222 return $httpurl;
2223}
2224
2225/**
8bd0068d 2226 * This function changes the <body> tag into a <div> tag since we
2227 * can't really have a body-within-body.
2228 *
2229 * @param $attary an array of attributes and values of <body>
2230 * @param $mailbox mailbox we're currently reading (for cid2http)
2231 * @param $message current message (for cid2http)
2232 * @param $id current message id (for cid2http)
2233 * @return a modified array of attributes to be set for <div>
2234 */
2dd879b8 2235function sq_body2div($attary, $mailbox, $message, $id){
b583c3e8 2236 $me = 'sq_body2div';
3d8371be 2237 $divattary = Array('class' => "'bodyclass'");
b583c3e8 2238 $text = '#000000';
c189a963 2239 $has_bgc_stl = $has_txt_stl = false;
b583c3e8 2240 $styledef = '';
691a2d25 2241 if (is_array($attary) && sizeof($attary) > 0){
2242 foreach ($attary as $attname=>$attvalue){
2243 $quotchar = substr($attvalue, 0, 1);
2244 $attvalue = str_replace($quotchar, "", $attvalue);
2245 switch ($attname){
3d8371be 2246 case 'background':
8bd0068d 2247 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
3d8371be 2248 $styledef .= "background-image: url('$attvalue'); ";
2249 break;
2250 case 'bgcolor':
c189a963 2251 $has_bgc_stl = true;
3d8371be 2252 $styledef .= "background-color: $attvalue; ";
2253 break;
2254 case 'text':
c189a963 2255 $has_txt_stl = true;
3d8371be 2256 $styledef .= "color: $attvalue; ";
2257 break;
691a2d25 2258 }
a3daaaf3 2259 }
c189a963 2260 // Outlook defines a white bgcolor and no text color. This can lead to
2261 // white text on a white bg with certain themes.
2262 if ($has_bgc_stl && !$has_txt_stl) {
2263 $styledef .= "color: $text; ";
2264 }
691a2d25 2265 if (strlen($styledef) > 0){
2266 $divattary{"style"} = "\"$styledef\"";
2267 }
2268 }
2269 return $divattary;
2270}
a3daaaf3 2271
691a2d25 2272/**
8bd0068d 2273 * This is the main function and the one you should actually be calling.
2274 * There are several variables you should be aware of an which need
2275 * special description.
2276 *
2277 * Since the description is quite lengthy, see it here:
2278 * http://linux.duke.edu/projects/mini/htmlfilter/
2279 *
2280 * @param $body the string with HTML you wish to filter
2281 * @param $tag_list see description above
2282 * @param $rm_tags_with_content see description above
2283 * @param $self_closing_tags see description above
2284 * @param $force_tag_closing see description above
2285 * @param $rm_attnames see description above
2286 * @param $bad_attvals see description above
2287 * @param $add_attr_to_tag see description above
2288 * @param $message message object
2289 * @param $id message id
2290 * @return sanitized html safe to show on your pages.
2291 */
da2415c1 2292function sq_sanitize($body,
8bd0068d 2293 $tag_list,
2294 $rm_tags_with_content,
2295 $self_closing_tags,
2296 $force_tag_closing,
2297 $rm_attnames,
2298 $bad_attvals,
2299 $add_attr_to_tag,
2300 $message,
2301 $id,
2302 $mailbox
2303 ){
b583c3e8 2304 $me = 'sq_sanitize';
7d06541f 2305 $rm_tags = array_shift($tag_list);
691a2d25 2306 /**
8bd0068d 2307 * Normalize rm_tags and rm_tags_with_content.
2308 */
7d06541f 2309 @array_walk($tag_list, 'sq_casenormalize');
691a2d25 2310 @array_walk($rm_tags_with_content, 'sq_casenormalize');
2311 @array_walk($self_closing_tags, 'sq_casenormalize');
2312 /**
8bd0068d 2313 * See if tag_list is of tags to remove or tags to allow.
2314 * false means remove these tags
2315 * true means allow these tags
2316 */
691a2d25 2317 $curpos = 0;
2318 $open_tags = Array();
2dd879b8 2319 $trusted = "\n<!-- begin sanitized html -->\n";
691a2d25 2320 $skip_content = false;
bb8d0799 2321 /**
8bd0068d 2322 * Take care of netscape's stupid javascript entities like
2323 * &{alert('boo')};
2324 */
bb8d0799 2325 $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
691a2d25 2326
7d06541f 2327 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
691a2d25 2328 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
2329 $free_content = substr($body, $curpos, $lt-$curpos);
2330 /**
8bd0068d 2331 * Take care of <style>
2332 */
7d06541f 2333 if ($tagname == "style" && $tagtype == 1){
da2415c1 2334 list($free_content, $curpos) =
e60a299a 2335 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
7d06541f 2336 if ($free_content != FALSE){
2337 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2338 $trusted .= $free_content;
2339 $trusted .= sq_tagprint($tagname, false, 2);
2340 }
2341 continue;
691a2d25 2342 }
2343 if ($skip_content == false){
2344 $trusted .= $free_content;
691a2d25 2345 }
2346 if ($tagname != FALSE){
2347 if ($tagtype == 2){
2348 if ($skip_content == $tagname){
2349 /**
8bd0068d 2350 * Got to the end of tag we needed to remove.
2351 */
691a2d25 2352 $tagname = false;
2353 $skip_content = false;
2354 } else {
2355 if ($skip_content == false){
c828931c 2356 if ($tagname == "body"){
2357 $tagname = "div";
2dd879b8 2358 }
da2415c1 2359 if (isset($open_tags{$tagname}) &&
8bd0068d 2360 $open_tags{$tagname} > 0){
2dd879b8 2361 $open_tags{$tagname}--;
691a2d25 2362 } else {
2dd879b8 2363 $tagname = false;
691a2d25 2364 }
691a2d25 2365 }
2366 }
2367 } else {
2368 /**
8bd0068d 2369 * $rm_tags_with_content
2370 */
691a2d25 2371 if ($skip_content == false){
2372 /**
8bd0068d 2373 * See if this is a self-closing type and change
2374 * tagtype appropriately.
2375 */
691a2d25 2376 if ($tagtype == 1
8bd0068d 2377 && in_array($tagname, $self_closing_tags)){
2dd879b8 2378 $tagtype = 3;
691a2d25 2379 }
2380 /**
8bd0068d 2381 * See if we should skip this tag and any content
2382 * inside it.
2383 */
691a2d25 2384 if ($tagtype == 1 &&
8bd0068d 2385 in_array($tagname, $rm_tags_with_content)){
691a2d25 2386 $skip_content = $tagname;
2387 } else {
da2415c1 2388 if (($rm_tags == false
8bd0068d 2389 && in_array($tagname, $tag_list)) ||
2390 ($rm_tags == true &&
2391 !in_array($tagname, $tag_list))){
691a2d25 2392 $tagname = false;
2393 } else {
2dd879b8 2394 /**
8bd0068d 2395 * Convert body into div.
2396 */
2dd879b8 2397 if ($tagname == "body"){
2398 $tagname = "div";
da2415c1 2399 $attary = sq_body2div($attary, $mailbox,
8bd0068d 2400 $message, $id);
2dd879b8 2401 }
691a2d25 2402 if ($tagtype == 1){
2403 if (isset($open_tags{$tagname})){
2404 $open_tags{$tagname}++;
2405 } else {
2406 $open_tags{$tagname}=1;
2407 }
2408 }
2409 /**
8bd0068d 2410 * This is where we run other checks.
2411 */
691a2d25 2412 if (is_array($attary) && sizeof($attary) > 0){
2413 $attary = sq_fixatts($tagname,
8bd0068d 2414 $attary,
2415 $rm_attnames,
2416 $bad_attvals,
2417 $add_attr_to_tag,
2418 $message,
2419 $id,
2420 $mailbox
2421 );
691a2d25 2422 }
2423 }
2424 }
691a2d25 2425 }
2426 }
2427 if ($tagname != false && $skip_content == false){
2428 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2429 }
691a2d25 2430 }
2431 $curpos = $gt+1;
a3daaaf3 2432 }
691a2d25 2433 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2434 if ($force_tag_closing == true){
2435 foreach ($open_tags as $tagname=>$opentimes){
2436 while ($opentimes > 0){
2437 $trusted .= '</' . $tagname . '>';
2438 $opentimes--;
2439 }
2440 }
2441 $trusted .= "\n";
2442 }
2443 $trusted .= "<!-- end sanitized html -->\n";
2444 return $trusted;
2445}
451f74a2 2446
691a2d25 2447/**
8bd0068d 2448 * This is a wrapper function to call html sanitizing routines.
2449 *
2450 * @param $body the body of the message
2451 * @param $id the id of the message
c189a963 2452
2453 * @param $message
2454 * @param $mailbox
2455 * @param boolean $take_mailto_links When TRUE, converts mailto: links
2456 * into internal SM compose links
2457 * (optional; default = TRUE)
8bd0068d 2458 * @return a string with html safe to display in the browser.
2459 */
c189a963 2460function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2461
202bcbcc 2462 // require_once(SM_PATH . 'functions/url_parser.php'); // for $MailTo_PReg_Match
c189a963 2463
691a2d25 2464 global $attachment_common_show_images, $view_unsafe_images,
8bd0068d 2465 $has_unsafe_images;
691a2d25 2466 /**
8bd0068d 2467 * Don't display attached images in HTML mode.
2b665f28 2468 *
d0187bd6 2469 * SB: why?
8bd0068d 2470 */
691a2d25 2471 $attachment_common_show_images = false;
2472 $tag_list = Array(
8bd0068d 2473 false,
2474 "object",
2475 "meta",
2476 "html",
2477 "head",
2478 "base",
2479 "link",
2480 "frame",
2481 "iframe",
2482 "plaintext",
2483 "marquee"
2484 );
691a2d25 2485
2486 $rm_tags_with_content = Array(
8bd0068d 2487 "script",
2488 "applet",
2489 "embed",
2490 "title",
2491 "frameset",
0493ed11 2492 "xmp",
8bd0068d 2493 "xml"
2494 );
691a2d25 2495
2496 $self_closing_tags = Array(
8bd0068d 2497 "img",
2498 "br",
2499 "hr",
2500 "input",
2501 "outbind"
2502 );
691a2d25 2503
2dd879b8 2504 $force_tag_closing = true;
691a2d25 2505
2506 $rm_attnames = Array(
8bd0068d 2507 "/.*/" =>
2508 Array(
2509 "/target/i",
2510 "/^on.*/i",
2511 "/^dynsrc/i",
2512 "/^data.*/i",
2513 "/^lowsrc.*/i"
2514 )
2515 );
691a2d25 2516
2517 $secremoveimg = "../images/" . _("sec_remove_eng.png");
2518 $bad_attvals = Array(
8bd0068d 2519 "/.*/" =>
691a2d25 2520 Array(
0a6ec9b5 2521 "/^src|background/i" =>
8bd0068d 2522 Array(
691a2d25 2523 Array(
8bd0068d 2524 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2525 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2526 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
691a2d25 2527 ),
8bd0068d 2528 Array(
2529 "\\1$secremoveimg\\2",
2530 "\\1$secremoveimg\\2",
2531 "\\1$secremoveimg\\2",
8bd0068d 2532 )
2533 ),
0a6ec9b5 2534 "/^href|action/i" =>
8bd0068d 2535 Array(
0a6ec9b5 2536 Array(
8bd0068d 2537 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2538 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2539 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
0a6ec9b5 2540 ),
691a2d25 2541 Array(
8bd0068d 2542 "\\1#\\1",
2543 "\\1#\\1",
2544 "\\1#\\1"
02474e43 2545 )
8bd0068d 2546 ),
2547 "/^style/i" =>
2548 Array(
2549 Array(
5db90261 2550 "/\/\*.*\*\//",
8bd0068d 2551 "/expression/i",
2552 "/binding/i",
2553 "/behaviou*r/i",
2554 "/include-source/i",
2555 "/position\s*:\s*absolute/i",
1d935bc2 2556 "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
8bd0068d 2557 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2558 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2559 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2560 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2561 ),
2562 Array(
5db90261 2563 "",
8bd0068d 2564 "idiocy",
2565 "idiocy",
2566 "idiocy",
2567 "idiocy",
567dc524 2568 "idiocy",
5b4884be 2569 "url",
8bd0068d 2570 "url(\\1#\\1)",
2571 "url(\\1#\\1)",
2572 "url(\\1#\\1)",
8bd0068d 2573 "\\1:url(\\2#\\3)"
2574 )
691a2d25 2575 )
8bd0068d 2576 )
691a2d25 2577 );
9ebace19 2578
d31f73f1 2579 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
2580 // images off by default.
afbf184c 2581 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET, FALSE);
9ebace19 2582
691a2d25 2583 if (!$view_unsafe_images){
2584 /**
8bd0068d 2585 * Remove any references to http/https if view_unsafe_images set
2586 * to false.
2587 */
02474e43 2588 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
8bd0068d 2589 '/^([\'\"])\s*https*:.*([\'\"])/si');
02474e43 2590 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
8bd0068d 2591 "\\1$secremoveimg\\1");
02474e43 2592 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
7ef0b415 2593 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
02474e43 2594 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
8bd0068d 2595 "url(\\1$secremoveimg\\1)");
691a2d25 2596 }
451f74a2 2597
691a2d25 2598 $add_attr_to_tag = Array(
8bd0068d 2599 "/^a$/i" =>
c25f2fbb 2600 Array('target'=>'"_blank"',
02474e43 2601 'title'=>'"'._("This external link will open in a new window").'"'
8bd0068d 2602 )
2603 );
da2415c1 2604 $trusted = sq_sanitize($body,
8bd0068d 2605 $tag_list,
2606 $rm_tags_with_content,
2607 $self_closing_tags,
2608 $force_tag_closing,
2609 $rm_attnames,
2610 $bad_attvals,
2611 $add_attr_to_tag,
2612 $message,
2613 $id,
2614 $mailbox
2615 );
567dc524 2616 if (strpos($trusted,$secremoveimg)){
691a2d25 2617 $has_unsafe_images = true;
da2415c1 2618 }
c189a963 2619
2620 // we want to parse mailto's in HTML output, change to SM compose links
2621 // this is a modified version of code from url_parser.php... but Marc is
2622 // right: we need a better filtering implementation; adding this randomly
2623 // here is not a great solution
2624 //
2625 if ($take_mailto_links) {
2626 // parseUrl($trusted); // this even parses URLs inside of tags... too aggressive
2627 global $MailTo_PReg_Match;
202bcbcc 2628 $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1) ;
c189a963 2629 if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2630 foreach ($regs[0] as $i => $mailto_before) {
2631 $mailto_params = $regs[10][$i];
2632 // get rid of any tailing quote since we have to add send_to to the end
2633 //
2634 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2635 $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2636 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2637 $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2638
2639 if ($regs[1][$i]) { //if there is an email addr before '?', we need to merge it with the params
2640 $to = 'to=' . $regs[1][$i];
2641 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
2642 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2643 else {
2644 if ($mailto_params) //already some params, append to them
2645 $mailto_params .= '&amp;' . $to;
2646 else
2647 $mailto_params .= '?' . $to;
2648 }
2649 }
2650
2651 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2652
2653 // we'll already have target=_blank, no need to allow comp_in_new
2654 // here (which would be a lot more work anyway)
2655 //
2656 global $compose_new_win;
2657 $temp_comp_in_new = $compose_new_win;
2658 $compose_new_win = 0;
2659 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2660 $compose_new_win = $temp_comp_in_new;
2661
2662 // remove <a href=" and anything after the next quote (we only
2663 // need the uri, not the link HTML) in compose uri
2664 //
2665 $comp_uri = substr($comp_uri, 9);
2666 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2667 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2668 }
2669 }
2670 }
2671
691a2d25 2672 return $trusted;
451f74a2 2673}
a4a70693 2674
da2415c1 2675/**
8bd0068d 2676 * function SendDownloadHeaders - send file to the browser
2677 *
2678 * Original Source: SM core src/download.php
2679 * moved here to make it available to other code, and separate
2680 * front end from back end functionality.
2681 *
2682 * @param string $type0 first half of mime type
2683 * @param string $type1 second half of mime type
2684 * @param string $filename filename to tell the browser for downloaded file
2685 * @param boolean $force whether to force the download dialog to pop
2686 * @param optional integer $filesize send the Content-Header and length to the browser
2687 * @return void
2688 */
02474e43 2689function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2690 global $languages, $squirrelmail_language;
cfffd60b 2691 $isIE = $isIE6plus = false;
02474e43 2692
2693 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2694
2695 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
8bd0068d 2696 strstr($HTTP_USER_AGENT, 'Opera') === false) {
cfffd60b 2697 $isIE = true;
02474e43 2698 }
2699
cfffd60b 2700 if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2701 ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2702 $isIE6plus = true;
02474e43 2703 }
2704
2705 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 2706 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
02474e43 2707 $filename =
8bd0068d 2708 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
02474e43 2709 } else {
2710 $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
2711 }
2712
2713 // A Pox on Microsoft and it's Internet Explorer!
2714 //
2715 // IE has lots of bugs with file downloads.
2716 // It also has problems with SSL. Both of these cause problems
2717 // for us in this function.
2718 //
2719 // See this article on Cache Control headers and SSL
2720 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2721 //
2722 // The best thing you can do for IE is to upgrade to the latest
2723 // version
2724 //set all the Cache Control Headers for IE
2725 if ($isIE) {
2726 $filename=rawurlencode($filename);
2727 header ("Pragma: public");
8bd0068d 2728 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
02474e43 2729 header ("Cache-Control: post-check=0, pre-check=0", false);
8bd0068d 2730 header ("Cache-Control: private");
02474e43 2731
2732 //set the inline header for IE, we'll add the attachment header later if we need it
2733 header ("Content-Disposition: inline; filename=$filename");
2734 }
2735
2736 if (!$force) {
2737 // Try to show in browser window
2738 header ("Content-Disposition: inline; filename=\"$filename\"");
2739 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2740 } else {
2741 // Try to pop up the "save as" box
2742
2743 // IE makes this hard. It pops up 2 save boxes, or none.
2744 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2745 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2746 // But, according to Microsoft, it is "RFC compliant but doesn't
2747 // take into account some deviations that allowed within the
2748 // specification." Doesn't that mean RFC non-compliant?
2749 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2750
2751 // all browsers need the application/octet-stream header for this
2752 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2753
2754 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2755 // Do not have quotes around filename, but that applied to
2756 // "attachment"... does it apply to inline too?
2757 header ("Content-Disposition: attachment; filename=\"$filename\"");
2758
cfffd60b 2759 if ($isIE && !$isIE6plus) {
02474e43 2760 // This combination seems to work mostly. IE 5.5 SP 1 has
2761 // known issues (see the Microsoft Knowledge Base)
2762
2763 // This works for most types, but doesn't work with Word files
2764 header ("Content-Type: application/download; name=\"$filename\"");
7e2ff844 2765 header ("Content-Type: application/force-download; name=\"$filename\"");
02474e43 2766 // These are spares, just in case. :-)
2767 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2768 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2769 //header("Content-Type: application/octet-stream; name=\"$filename\"");
7e2ff844 2770 } else if ($isIE) {
2771 // This is to prevent IE for MIME sniffing and auto open a file in IE
2772 header ("Content-Type: application/force-download; name=\"$filename\"");
02474e43 2773 } else {
2774 // another application/octet-stream forces download for Netscape
2775 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2776 }
2777 }
2778
2779 //send the content-length header if the calling function provides it
2780 if ($filesize > 0) {
2781 header("Content-Length: $filesize");
2782 }
07c49f57 2783
8d863f64 2784} // end fn SendDownloadHeaders
Unnamed repository; edit this file 'description' to name the repository.