projects / squirrelmail.git / blame
? search:
summary | shortlog | log | commit | commitdiff | tree
blob | blame (incremental) | history | HEAD
Viewing unsafe images is a core functionality, so I remove these comments.
[squirrelmail.git] / functions / mime.php
CommitLineData
59177427 1<?php
2ba13803 2
35586184 3/**
8bd0068d 4 * mime.php
5 *
8bd0068d 6 * This contains the functions necessary to detect and decode MIME
7 * messages.
8 *
4b5049de 9 * @copyright &copy; 1999-2007 The SquirrelMail Project Team
4b4abf93 10 * @license http://opensource.org/licenses/gpl-license.php GNU Public License
8bd0068d 11 * @version $Id$
12 * @package squirrelmail
13 */
b74ba498 14
202bcbcc 15/**
16 * dependency information
17 functions dependency
18 mime_structure
19 class/mime/Message.class.php
20 Message::parseStructure
21 functions/page_header.php
22 displayPageHeader
23 functions/display_messages.php
24 plain_error_message
25 mime_fetch_body
26 functions/imap_general.php
27 sqimap_run_command
28 mime_print_body_lines
29
30
31
32functions/imap.php
33functions/attachment_common.php
34functions/display_messages.php
35
36magicHtml => url_parser
37translateText => url_parser
38
39*/
40
8beafbbc 41
7c7b74b3 42/* -------------------------------------------------------------------------- */
43/* MIME DECODING */
44/* -------------------------------------------------------------------------- */
b74ba498 45
d6c32258 46/**
8bd0068d 47 * Get the MIME structure
48 *
49 * This function gets the structure of a message and stores it in the "message" class.
50 * It will return this object for use with all relevant header information and
51 * fully parsed into the standard "message" object format.
52 */
a4a70693 53function mime_structure ($bodystructure, $flags=array()) {
c9d78ab4 54
3d8371be 55 /* Isolate the body structure and remove beginning and end parenthesis. */
a4a70693 56 $read = trim(substr ($bodystructure, strpos(strtolower($bodystructure), 'bodystructure') + 13));
451f74a2 57 $read = trim(substr ($read, 0, -1));
22efa9fb 58 $i = 0;
59 $msg = Message::parseStructure($read,$i);
2b665f28 60
9de42168 61 if (!is_object($msg)) {
3d8371be 62 global $color, $mailbox;
a48eba8f 63 displayPageHeader( $color, $mailbox );
5e8de8b6 64 $errormessage = _("SquirrelMail could not decode the bodystructure of the message");
8bd0068d 65 $errormessage .= '<br />'._("The bodystructure provided by your IMAP server:").'<br /><br />';
472e7acb 66 $errormessage .= '<pre>' . htmlspecialchars($read) . '</pre>';
ce8c6f42 67 plain_error_message( $errormessage );
3d8371be 68 echo '</body></html>';
9de42168 69 exit;
70 }
a4a70693 71 if (count($flags)) {
7a9e9c89 72 foreach ($flags as $flag) {
e1115979 73//FIXME: please document why it is we have to check the first char of the flag but we then go ahead and do a full string comparison anyway. Is this a speed enhancement? If not, let's keep it simple and just compare the full string and forget the switch block.
7a9e9c89 74 $char = strtoupper($flag{1});
75 switch ($char) {
3d8371be 76 case 'S':
77 if (strtolower($flag) == '\\seen') {
78 $msg->is_seen = true;
79 }
80 break;
81 case 'A':
82 if (strtolower($flag) == '\\answered') {
83 $msg->is_answered = true;
84 }
85 break;
86 case 'D':
87 if (strtolower($flag) == '\\deleted') {
88 $msg->is_deleted = true;
89 }
90 break;
91 case 'F':
92 if (strtolower($flag) == '\\flagged') {
93 $msg->is_flagged = true;
94 }
95 break;
96 case 'M':
97 if (strtolower($flag) == '$mdnsent') {
98 $msg->is_mdnsent = true;
99 }
100 break;
101 default:
102 break;
7a9e9c89 103 }
104 }
451f74a2 105 }
7a9e9c89 106 // listEntities($msg);
3d8371be 107 return $msg;
451f74a2 108}
b74ba498 109
22efa9fb 110
111
3d8371be 112/* This starts the parsing of a particular structure. It is called recursively,
8bd0068d 113 * so it can be passed different structures. It returns an object of type
114 * $message.
115 * First, it checks to see if it is a multipart message. If it is, then it
116 * handles that as it sees is necessary. If it is just a regular entity,
117 * then it parses it and adds the necessary header information (by calling out
118 * to mime_get_elements()
119 */
451f74a2 120
4d592352 121function mime_fetch_body($imap_stream, $id, $ent_id=1, $fetch_size=0) {
3d8371be 122 /* Do a bit of error correction. If we couldn't find the entity id, just guess
8bd0068d 123 * that it is the first one. That is usually the case anyway.
124 */
7c7b74b3 125
09a4bde3 126 if (!$ent_id) {
08b7f7cc 127 $cmd = "FETCH $id BODY[]";
1035e159 128 } else {
08b7f7cc 129 $cmd = "FETCH $id BODY[$ent_id]";
09a4bde3 130 }
3d8371be 131
4d592352 132 if ($fetch_size!=0) $cmd .= "<0.$fetch_size>";
da2415c1 133
6201339c 134 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, TRUE);
77b88425 135 do {
3d8371be 136 $topline = trim(array_shift($data));
137 } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
a4a70693 138
451f74a2 139 $wholemessage = implode('', $data);
140 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
3d8371be 141 $ret = substr($wholemessage, 0, $regs[1]);
142 /* There is some information in the content info header that could be important
8bd0068d 143 * in order to parse html messages. Let's get them here.
144 */
0600bdf1 145// if ($ret{0} == '<') {
6201339c 146// $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, TRUE);
0600bdf1 147// }
451f74a2 148 } else if (ereg('"([^"]*)"', $topline, $regs)) {
149 $ret = $regs[1];
06bcb9c3 150 } else if ((stristr($topline, 'nil') !== false) && (empty($wholemessage))) {
151 $ret = $wholemessage;
451f74a2 152 } else {
153 global $where, $what, $mailbox, $passed_id, $startMessage;
3d8371be 154 $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
451f74a2 155 if (isset($where) && isset($what)) {
3d8371be 156 $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
a3daaaf3 157 } else {
3d8371be 158 $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
451f74a2 159 }
e5ea9327 160 $par .= '&amp;response=' . urlencode($response) .
8bd0068d 161 '&amp;message=' . urlencode($message) .
162 '&amp;topline=' . urlencode($topline);
a019eeb8 163
8bd0068d 164 echo '<tt><br />' .
02474e43 165 '<table width="80%"><tr>' .
166 '<tr><td colspan="2">' .
167 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
168 '</td></tr>' .
169 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
170 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
171 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
172 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
173 "</table><br /></tt></font><hr />";
346817d4 174
6201339c 175 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, TRUE);
451f74a2 176 array_shift($data);
177 $wholemessage = implode('', $data);
a019eeb8 178
346817d4 179 $ret = $wholemessage;
a3daaaf3 180 }
3d8371be 181 return $ret;
451f74a2 182}
d4467150 183
6cc670de 184function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding, $rStream='php://stdout') {
1035e159 185
3d8371be 186 /* Don't kill the connection if the browser is over a dialup
8bd0068d 187 * and it would take over 30 seconds to download it.
188 * Don't call set_time_limit in safe mode.
189 */
b7206e1d 190
3d8371be 191 if (!ini_get('safe_mode')) {
b7206e1d 192 set_time_limit(0);
193 }
7c7b74b3 194 /* in case of base64 encoded attachments, do not buffer them.
8bd0068d 195 Instead, echo the decoded attachment directly to screen */
7c7b74b3 196 if (strtolower($encoding) == 'base64') {
197 if (!$ent_id) {
7591f143 198 $query = "FETCH $id BODY[]";
7c7b74b3 199 } else {
7591f143 200 $query = "FETCH $id BODY[$ent_id]";
7c7b74b3 201 }
7591f143 202 sqimap_run_command($imap_stream,$query,true,$response,$message,TRUE,'sqimap_base64_decode',$rStream,true);
1d142b8d 203 } else {
7591f143 204 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
205 if (is_resource($rStream)) {
206 fputs($rStream,decodeBody($body,$encoding));
207 } else {
208 echo decodeBody($body, $encoding);
209 }
1d142b8d 210 }
346817d4 211
da2415c1 212 /*
8bd0068d 213 TODO, use the same method for quoted printable.
214 However, I assume that quoted printable attachments aren't that large
215 so the performancegain / memory usage drop will be minimal.
216 If we decide to add that then we need to adapt sqimap_fread because
217 we need to split te result on \n and fread doesn't stop at \n. That
218 means we also should provide $results from sqimap_fread (by ref) to
219 te function and set $no_return to false. The $filter function for
220 quoted printable should handle unsetting of $results.
221 */
da2415c1 222 /*
8bd0068d 223 TODO 2: find out how we write to the output stream php://stdout. fwrite
224 doesn't work because 'php://stdout isn't a stream.
225 */
7c7b74b3 226
5d9c6f73 227 return;
451f74a2 228}
beb9e459 229
451f74a2 230/* -[ END MIME DECODING ]----------------------------------------------------------- */
d4467150 231
3d8371be 232/* This is here for debugging purposes. It will print out a list
8bd0068d 233 * of all the entity IDs that are in the $message object.
234 */
451f74a2 235function listEntities ($message) {
3d8371be 236 if ($message) {
3c621ba1 237 echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br />';
3d8371be 238 for ($i = 0; isset($message->entities[$i]); $i++) {
239 echo "$i : ";
240 $msg = listEntities($message->entities[$i]);
241
242 if ($msg) {
243 echo "return: ";
244 return $msg;
245 }
246 }
a4a70693 247 }
451f74a2 248}
f0c4dc12 249
f792c641 250function getPriorityStr($priority) {
3d8371be 251 $priority_level = substr($priority,0,1);
252
253 switch($priority_level) {
254 /* Check for a higher then normal priority. */
255 case '1':
256 case '2':
257 $priority_string = _("High");
258 break;
259
260 /* Check for a lower then normal priority. */
261 case '4':
262 case '5':
263 $priority_string = _("Low");
264 break;
265
266 /* Check for a normal priority. */
267 case '3':
268 default:
269 $priority_level = '3';
270 $priority_string = _("Normal");
271 break;
272
273 }
274 return $priority_string;
f792c641 275}
276
451f74a2 277/* returns a $message object for a particular entity id */
278function getEntity ($message, $ent_id) {
a4a70693 279 return $message->getEntity($ent_id);
451f74a2 280}
8beafbbc 281
3d8371be 282/* translateText
8bd0068d 283 * Extracted from strings.php 23/03/2002
284 */
da4c66e8 285
286function translateText(&$body, $wrap_at, $charset) {
3d8371be 287 global $where, $what; /* from searching */
288 global $color; /* color theme */
da4c66e8 289
202bcbcc 290 // require_once(SM_PATH . 'functions/url_parser.php');
da4c66e8 291
292 $body_ary = explode("\n", $body);
da4c66e8 293 for ($i=0; $i < count($body_ary); $i++) {
294 $line = $body_ary[$i];
295 if (strlen($line) - 2 >= $wrap_at) {
c7aff938 296 sqWordWrap($line, $wrap_at, $charset);
da4c66e8 297 }
298 $line = charset_decode($charset, $line);
299 $line = str_replace("\t", ' ', $line);
300
301 parseUrl ($line);
302
3d8371be 303 $quotes = 0;
da4c66e8 304 $pos = 0;
3d8371be 305 $j = strlen($line);
da4c66e8 306
3d8371be 307 while ($pos < $j) {
da4c66e8 308 if ($line[$pos] == ' ') {
3d8371be 309 $pos++;
da4c66e8 310 } else if (strpos($line, '&gt;', $pos) === $pos) {
311 $pos += 4;
3d8371be 312 $quotes++;
da4c66e8 313 } else {
314 break;
315 }
316 }
3d8371be 317
83c94382 318 if ($quotes % 2) {
d0814c02 319 $line = '<span class="quote1">' . $line . '</span>';
4c25967c 320 } elseif ($quotes) {
d0814c02 321 $line = '<span class="quote2">' . $line . '</span>';
da4c66e8 322 }
3d8371be 323
da4c66e8 324 $body_ary[$i] = $line;
325 }
326 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
327}
328
a2bfcbce 329/**
da1b55ad 330 * This returns a parsed string called $body. That string can then
331 * be displayed as the actual message in the HTML. It contains
332 * everything needed, including HTML Tags, Attachments at the
333 * bottom, etc.
f8a1ed5a 334 *
da1b55ad 335 * Since 1.2.0 function uses message_body hook.
336 * Till 1.3.0 function included output of formatAttachments().
337 *
338 * @param resource $imap_stream imap connection resource
339 * @param object $message squirrelmail message object
340 * @param array $color squirrelmail color theme array
341 * @param integer $wrap_at number of characters per line
342 * @param string $ent_num (since 1.3.0) message part id
343 * @param integer $id (since 1.3.0) message id
344 * @param string $mailbox (since 1.3.0) imap folder name
41f701c1 345 * @param boolean $clean (since 1.5.1) Do not output stuff that's irrelevant for the printable version.
da1b55ad 346 * @return string html formated message text
347 */
a2bfcbce 348function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX', $clean=FALSE) {
3d8371be 349 /* This if statement checks for the entity to show as the
8bd0068d 350 * primary message. To add more of them, just put them in the
351 * order that is their priority.
352 */
ce68b76b 353 global $startMessage, $languages, $squirrelmail_language,
40a34e57 354 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id,
14c85e39 355 $use_iframe, $iframe_height, $download_and_unsafe_link,
955bfc8f 356 $download_href, $unsafe_image_toggle_href, $unsafe_image_toggle_text,
551c7b53 357 $oTemplate, $nbsp;
2c25d36a 358
359 // workaround for not updated config.php
360 if (! isset($use_iframe)) $use_iframe = false;
77bfbd2e 361
9ebace19 362 // If there's no "view_unsafe_images" setting in the user's preferences,
363 // turn unsafe images off by default.
5262d9a6 364 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
365 $view_unsafe_images = false;
77bfbd2e 366 }
d03c24f4 367
cc34b00d 368 $body = '';
23bcec6f 369 $urlmailbox = urlencode($mailbox);
451f74a2 370 $body_message = getEntity($message, $ent_num);
371 if (($body_message->header->type0 == 'text') ||
8bd0068d 372 ($body_message->header->type0 == 'rfc822')) {
3d8371be 373 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
451f74a2 374 $body = decodeBody($body, $body_message->header->encoding);
e842b215 375
376 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 377 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
e842b215 378 if (mb_detect_encoding($body) != 'ASCII') {
33a55f5a 379 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
e842b215 380 }
381 }
d849b570 382
383 /* As of 1.5.2, $body is passed (and modified) by reference */
384 do_hook('message_body', $body);
23bcec6f 385
3d8371be 386 /* If there are other types that shouldn't be formatted, add
8bd0068d 387 * them here.
388 */
3d8371be 389
451f74a2 390 if ($body_message->header->type1 == 'html') {
3d8371be 391 if ($show_html_default <> 1) {
85015544 392 $entity_conv = array('&nbsp;' => ' ',
8bd0068d 393 '<p>' => "\n",
394 '<P>' => "\n",
395 '<br>' => "\n",
396 '<BR>' => "\n",
397 '<br />' => "\n",
398 '<BR />' => "\n",
399 '&gt;' => '>',
400 '&lt;' => '<');
85015544 401 $body = strtr($body, $entity_conv);
3d8371be 402 $body = strip_tags($body);
85015544 403 $body = trim($body);
404 translateText($body, $wrap_at,
8bd0068d 405 $body_message->header->getParameter('charset'));
2c25d36a 406 } elseif ($use_iframe && ! $clean) {
407 // $clean is used to remove iframe in printable view.
408
84410f31 409 /**
410 * If we don't add html message between iframe tags,
411 * we must detect unsafe images and modify $has_unsafe_images.
f8a1ed5a 412 */
758a7889 413 $html_body = magicHTML($body, $id, $message, $mailbox);
b6c52e61 414 // Convert character set in order to display html mails in different character set
415 $html_body = charset_decode($body_message->header->getParameter('charset'),$html_body,false,true);
84410f31 416
2c25d36a 417 // creating iframe url
418 $iframeurl=sqm_baseuri().'src/view_html.php?'
f8a1ed5a 419 . 'mailbox=' . $urlmailbox
2c25d36a 420 . '&amp;passed_id=' . $id
421 . '&amp;ent_id=' . $ent_num
422 . '&amp;view_unsafe_images=' . (int) $view_unsafe_images;
423
79d58d4c 424 global $oTemplate;
425 $oTemplate->assign('iframe_url', $iframeurl);
51554708 426 $oTemplate->assign('iframe_height', $iframe_height);
79d58d4c 427 $oTemplate->assign('html_body', $html_body);
2b665f28 428
79d58d4c 429 $body = $oTemplate->fetch('read_html_iframe.tpl');
a3daaaf3 430 } else {
2c25d36a 431 // old way of html rendering
b6c52e61 432 /**
758a7889 433 * convert character set. charset_decode does not remove html special chars
b6c52e61 434 * applied by magicHTML functions and does not sanitize them second time if
758a7889 435 * fourth argument is true.
436 */
567dc524 437 $charset = $body_message->header->getParameter('charset');
438 if (!empty($charset)) {
439 $body = charset_decode($charset,$body,false,true);
440 }
441 $body = magicHTML($body, $id, $message, $mailbox);
a3daaaf3 442 }
451f74a2 443 } else {
3d8371be 444 translateText($body, $wrap_at,
8bd0068d 445 $body_message->header->getParameter('charset'));
451f74a2 446 }
a2bfcbce 447
448 // if this is the clean display (i.e. printer friendly), stop here.
449 if ( $clean ) {
450 return $body;
451 }
452
9ebace19 453 /*
454 * Previously the links for downloading and unsafe images were printed
455 * under the mail. By putting the links in a global variable we can
456 * print it in the toolbar where it belongs. Since the original code was
457 * in this place it's left here. It might be possible to move it to some
458 * other place if that makes sense. The possibility to do so has not
459 * been evaluated yet.
460 */
461
462 // Initialize the global variable to an empty string.
463 // FIXME: To have $download_and_unsafe_link as a global variable might not be needed since the use of separate variables ($download_href, $unsafe_image_toggle_href, and $unsafe_image_toggle_text) for the templates was introduced.
40a34e57 464 $download_and_unsafe_link = '';
465
9ebace19 466 // Prepare and build a link for downloading the mail.
83cf04bd 467 $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
8bd0068d 468 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
469 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
08b7f7cc 470 if (isset($passed_ent_id)) {
471 $link .= '&amp;passed_ent_id='.$passed_ent_id;
472 }
14c85e39 473 $download_href = SM_PATH . 'src/download.php?absolute_dl=true&amp;' . $link;
9ebace19 474
475 // Always add the link for downloading the mail as a file to the global
476 // variable.
955bfc8f 477 $download_and_unsafe_link .= "$nbsp|$nbsp"
478 . create_hyperlink($download_href, _("Download this as a file"));
9ebace19 479
480 // Find out the right text to use in the link depending on the
481 // circumstances. If the unsafe images are displayed the link should
482 // hide them, if they aren't displayed the link should only appear if
483 // the mail really contains unsafe images.
7aad7b77 484 if ($view_unsafe_images) {
23f617b8 485 $text = _("Hide Unsafe Images");
7aad7b77 486 } else {
08b7f7cc 487 if (isset($has_unsafe_images) && $has_unsafe_images) {
488 $link .= '&amp;view_unsafe_images=1';
489 $text = _("View Unsafe Images");
490 } else {
491 $text = '';
492 }
3d8371be 493 }
9ebace19 494
495 // Only create a link for unsafe images if there's need for one. If so:
496 // add it to the global variable.
83cf04bd 497 if($text != '') {
14c85e39 498 $unsafe_image_toggle_href = SM_PATH . 'src/read_body.php?'.$link;
499 $unsafe_image_toggle_text = $text;
955bfc8f 500 $download_and_unsafe_link .= "$nbsp|$nbsp"
501 . create_hyperlink($unsafe_image_toggle_href, $text);
83cf04bd 502 }
3d8371be 503 }
504 return $body;
451f74a2 505}
b74ba498 506
da1b55ad 507/**
d67f519a 508 * Generate attachments array for passing to templates. Separated from
509 * formatAttachments() below so that the same array can be given to the
510 * print-friendly version.
2b665f28 511 *
d67f519a 512 * @since 1.5.2
da1b55ad 513 * @param object $message SquirrelMail message object
514 * @param array $exclude_id message parts that are not attachments.
515 * @param string $mailbox mailbox name
516 * @param integer $id message id
da1b55ad 517 */
d67f519a 518function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
519 global $where, $what, $startMessage, $color, $passed_ent_id, $base_uri;
451f74a2 520
23bcec6f 521 $att_ar = $message->getAttachments($exclude_id);
23bcec6f 522 $urlMailbox = urlencode($mailbox);
523
d67f519a 524 $attachments = array();
23bcec6f 525 foreach ($att_ar as $att) {
fdc9d9b5 526 $ent = $att->entity_id;
2e25760a 527 $header = $att->header;
f0c4dc12 528 $type0 = strtolower($header->type0);
529 $type1 = strtolower($header->type1);
2e25760a 530 $name = '';
d0187bd6 531 $links = array();
21dab2dc 532 $links['download link']['text'] = _("Download");
202bcbcc 533 $links['download link']['href'] = $base_uri .
8bd0068d 534 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
2b665f28 535
2e25760a 536 if ($type0 =='message' && $type1 == 'rfc822') {
202bcbcc 537 $default_page = $base_uri . 'src/read_body.php';
2e25760a 538 $rfc822_header = $att->rfc822_header;
098ea084 539 $filename = $rfc822_header->subject;
6cc08d8b 540 if (trim( $filename ) == '') {
541 $filename = 'untitled-[' . $ent . ']' ;
08b7f7cc 542 }
2e25760a 543 $from_o = $rfc822_header->from;
544 if (is_object($from_o)) {
04ea844e 545 $from_name = decodeHeader($from_o->getAddress(false));
7e697748 546 } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
547 // something weird happens when a digest message is opened and you return to the digest
548 // now the from object is part of an array. Probably the parseHeader call overwrites the info
549 // retrieved from the bodystructure in a different way. We need to fix this later.
550 // possible starting point, do not fetch header we already have and inspect how
551 // the rfc822_header object behaves.
552 $from_name = decodeHeader($from_o[0]->getAddress(false));
2e25760a 553 } else {
554 $from_name = _("Unknown sender");
f0c4dc12 555 }
d0187bd6 556 $description = _("From").': '.$from_name;
23bcec6f 557 } else {
202bcbcc 558 $default_page = $base_uri . 'src/download.php';
02474e43 559 $filename = $att->getFilename();
f810c0b2 560 if ($header->description) {
098ea084 561 $description = decodeHeader($header->description);
f810c0b2 562 } else {
3d8371be 563 $description = '';
564 }
2e25760a 565 }
566
567 $display_filename = $filename;
3d8371be 568 if (isset($passed_ent_id)) {
569 $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
570 } else {
571 $passed_ent_id_link = '';
572 }
573 $defaultlink = $default_page . "?startMessage=$startMessage"
8bd0068d 574 . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
575 . '&amp;ent_id='.$ent.$passed_ent_id_link;
2e25760a 576 if ($where && $what) {
8bd0068d 577 $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
2e25760a 578 }
7e2ff844 579 // IE does make use of mime content sniffing. Forcing a download
580 // prohibit execution of XSS inside an application/octet-stream attachment
581 if ($type0 == 'application' && $type1 == 'octet-stream') {
582 $defaultlink .= '&amp;absolute_dl=true';
583 }
2b665f28 584
3d8371be 585 /* This executes the attachment hook with a specific MIME-type.
8bd0068d 586 * If that doesn't have results, it tries if there's a rule
9b94c54d 587 * for a more generic type. Finally, a hook for ALL attachment
588 * types is run as well.
8bd0068d 589 */
8dca4d22 590 // First remember the default link.
591 $defaultlink_orig = $defaultlink;
592
d849b570 593 /* The API for this hook has changed as of 1.5.2 so that all plugin
594 arguments are passed in an array instead of each their own plugin
595 argument, and arguments are passed by reference, so instead of
596 returning any changes, changes should simply be made to the original
597 arguments themselves. */
9c3b2d22 598 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
599 &$defaultlink, &$display_filename, &$where, &$what);
600 do_hook("attachment $type0/$type1", $temp);
8dca4d22 601 if(count($links) <= 1 && $defaultlink == $defaultlink_orig) {
d849b570 602 /* The API for this hook has changed as of 1.5.2 so that all plugin
603 arguments are passed in an array instead of each their own plugin
604 argument, and arguments are passed by reference, so instead of
605 returning any changes, changes should simply be made to the original
606 arguments themselves. */
9c3b2d22 607 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
608 &$defaultlink, &$display_filename, &$where, &$what);
609 do_hook("attachment $type0/*", $temp);
2e25760a 610 }
d849b570 611 /* The API for this hook has changed as of 1.5.2 so that all plugin
612 arguments are passed in an array instead of each their own plugin
613 argument, and arguments are passed by reference, so instead of
614 returning any changes, changes should simply be made to the original
615 arguments themselves. */
9c3b2d22 616 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
617 &$defaultlink, &$display_filename, &$where, &$what);
8dca4d22 618 // Do not let a generic plugin change the default link if a more
619 // specialized one already did it...
620 if ($defaultlink != $defaultlink_orig) {
621 $dummy = '';
622 $temp[5] = &$dummy;
623 }
9c3b2d22 624 do_hook("attachment */*", $temp);
77b88425 625
d67f519a 626 $this_attachment = array();
627 $this_attachment['Name'] = decodeHeader($display_filename);
628 $this_attachment['Description'] = $description;
629 $this_attachment['DefaultHREF'] = $defaultlink;
630 $this_attachment['DownloadHREF'] = $links['download link']['href'];
631 $this_attachment['ViewHREF'] = isset($links['attachment_common']) ? $links['attachment_common']['href'] : '';
632 $this_attachment['Size'] = $header->size;
633 $this_attachment['ContentType'] = htmlspecialchars($type0 .'/'. $type1);
634 $this_attachment['OtherLinks'] = array();
3d8371be 635 foreach ($links as $val) {
d0187bd6 636 if ($val['text']==_("Download") || $val['text'] == _("View"))
637 continue;
638 if (empty($val['text']) && empty($val['extra']))
639 continue;
2b665f28 640
d67f519a 641 $temp = array();
642 $temp['HREF'] = $val['href'];
643 $temp['Text'] = (empty($val['text']) ? '' : $val['text']) . (empty($val['extra']) ? '' : $val['extra']);
644 $this_attachment['OtherLinks'][] = $temp;
2e25760a 645 }
d67f519a 646 $attachments[] = $this_attachment;
2b665f28 647
3d8371be 648 unset($links);
2e25760a 649 }
2b665f28 650
d67f519a 651 return $attachments;
652}
653
654/**
655 * Displays attachment links and information
656 *
657 * Since 1.3.0 function is not included in formatBody() call.
658 *
659 * Since 1.0.2 uses attachment $type0/$type1 hook.
660 * Since 1.2.5 uses attachment $type0/* hook.
661 * Since 1.5.0 uses attachments_bottom hook.
662 * Since 1.5.2 uses templates and does *not* return a value.
663 *
664 * @param object $message SquirrelMail message object
665 * @param array $exclude_id message parts that are not attachments.
666 * @param string $mailbox mailbox name
667 * @param integer $id message id
668 */
669function formatAttachments($message, $exclude_id, $mailbox, $id) {
670 global $oTemplate;
2b665f28 671
d67f519a 672 $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
d0187bd6 673
674 $oTemplate->assign('attachments', $attach);
675 $oTemplate->display('read_attachments.tpl');
451f74a2 676}
b74ba498 677
7c7b74b3 678function sqimap_base64_decode(&$string) {
7c0ec1d8 679
b17a8968 680 // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
7c0ec1d8 681 // fly decoding (to reduce memory usage) you have to check if the
682 // data has incomplete pairs
683
b17a8968 684 // Remove the noise in order to check if the 4 bytes pairs are complete
7c0ec1d8 685 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
686
42ce44f8 687 $sStringRem = '';
7c0ec1d8 688 $iMod = strlen($string) % 4;
689 if ($iMod) {
690 $sStringRem = substr($string,-$iMod);
b17a8968 691 // Check if $sStringRem contains padding characters
7c0ec1d8 692 if (substr($sStringRem,-1) != '=') {
693 $string = substr($string,0,-$iMod);
694 } else {
695 $sStringRem = '';
696 }
697 }
7c7b74b3 698 $string = base64_decode($string);
7c0ec1d8 699 return $sStringRem;
7c7b74b3 700}
701
fdf7cef1 702/**
703 * Decodes encoded message body
704 *
705 * This function decodes the body depending on the encoding type.
706 * Currently quoted-printable and base64 encodings are supported.
707 * decode_body hook was added to this function in 1.4.2/1.5.0
708 * @param string $body encoded message body
709 * @param string $encoding used encoding
710 * @return string decoded string
711 * @since 1.0
712 */
451f74a2 713function decodeBody($body, $encoding) {
83be314a 714
b583c3e8 715 $body = str_replace("\r\n", "\n", $body);
716 $encoding = strtolower($encoding);
3d8371be 717
d849b570 718 $encoding_handler = do_hook('decode_body', $encoding);
5166f86a 719
720
721 // plugins get first shot at decoding the body
722 //
723 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
724 $body = $encoding_handler('decode', $body);
725
fdf7cef1 726 } elseif ($encoding == 'quoted-printable' ||
8bd0068d 727 $encoding == 'quoted_printable') {
fdf7cef1 728 /**
729 * quoted_printable_decode() function is broken in older
730 * php versions. Text with \r\n decoding was fixed only
758a7889 731 * in php 4.3.0. Minimal code requirement 4.0.4 +
fdf7cef1 732 * str_replace("\r\n", "\n", $body); call.
733 */
b583c3e8 734 $body = quoted_printable_decode($body);
fdf7cef1 735 } elseif ($encoding == 'base64') {
b583c3e8 736 $body = base64_decode($body);
737 }
3d8371be 738
b583c3e8 739 // All other encodings are returned raw.
3d8371be 740 return $body;
451f74a2 741}
742
9f7f68c3 743/**
8bd0068d 744 * Decodes headers
745 *
e1115979 746 * This function decodes strings that are encoded according to
8bd0068d 747 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
748 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
749 *
750 * @param string $string header string that has to be made readable
751 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
bcae324b 752 * @param boolean $htmlsafe preserve spaces and sanitize html special characters. defaults to true
8bd0068d 753 * @param boolean $decide decide if string can be utfencoded. defaults to false
754 * @return string decoded header string
755 */
bcae324b 756function decodeHeader ($string, $utfencode=true,$htmlsafe=true,$decide=false) {
d6f584fc 757 global $languages, $squirrelmail_language,$default_charset;
79e07c7e 758 if (is_array($string)) {
759 $string = implode("\n", $string);
760 }
da2415c1 761
10dec454 762 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 763 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
33a55f5a 764 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
08b7f7cc 765 // Do we need to return at this point?
766 // return $string;
83be314a 767 }
79e07c7e 768 $i = 0;
08b7f7cc 769 $iLastMatch = -2;
db65b6b0 770 $encoded = true;
0a06275a 771
098ea084 772 $aString = explode(' ',$string);
08b7f7cc 773 $ret = '';
098ea084 774 foreach ($aString as $chunk) {
358a78a1 775 if ($encoded && $chunk === '') {
08b7f7cc 776 continue;
358a78a1 777 } elseif ($chunk === '') {
08b7f7cc 778 $ret .= ' ';
779 continue;
780 }
098ea084 781 $encoded = false;
08b7f7cc 782 /* if encoded words are not separated by a linear-space-white we still catch them */
783 $j = $i-1;
7e6ca3e8 784
08b7f7cc 785 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
786 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
787 if ($iLastMatch !== $j) {
bcae324b 788 if ($htmlsafe) {
9f7f68c3 789 $ret .= '&#32;';
08b7f7cc 790 } else {
791 $ret .= ' ';
792 }
793 }
794 $iLastMatch = $i;
795 $j = $i;
bcae324b 796 if ($htmlsafe) {
cb718de0 797 $ret .= htmlspecialchars($res[1]);
798 } else {
799 $ret .= $res[1];
800 }
098ea084 801 $encoding = ucfirst($res[3]);
d6f584fc 802
803 /* decide about valid decoding */
804 if ($decide && is_conversion_safe($res[2])) {
8bd0068d 805 $utfencode=true;
806 $can_be_encoded=true;
d6f584fc 807 } else {
8bd0068d 808 $can_be_encoded=false;
d6f584fc 809 }
098ea084 810 switch ($encoding)
811 {
8bd0068d 812 case 'B':
813 $replace = base64_decode($res[4]);
814 if ($utfencode) {
815 if ($can_be_encoded) {
816 /* convert string to different charset,
817 * if functions asks for it (usually in compose)
818 */
bcae324b 819 $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsafe);
8bd0068d 820 } else {
821 // convert string to html codes in order to display it
822 $ret .= charset_decode($res[2],$replace);
823 }
fab65ca9 824 } else {
bcae324b 825 if ($htmlsafe) {
8bd0068d 826 $replace = htmlspecialchars($replace);
827 }
828 $ret.= $replace;
fab65ca9 829 }
8bd0068d 830 break;
831 case 'Q':
832 $replace = str_replace('_', ' ', $res[4]);
833 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
834 $replace);
835 if ($utfencode) {
836 if ($can_be_encoded) {
837 /* convert string to different charset,
838 * if functions asks for it (usually in compose)
839 */
bcae324b 840 $replace = charset_convert($res[2], $replace,$default_charset,$htmlsafe);
8bd0068d 841 } else {
842 // convert string to html codes in order to display it
843 $replace = charset_decode($res[2], $replace);
844 }
845 } else {
bcae324b 846 if ($htmlsafe) {
8bd0068d 847 $replace = htmlspecialchars($replace);
848 }
098ea084 849 }
8bd0068d 850 $ret .= $replace;
851 break;
852 default:
853 break;
79e07c7e 854 }
098ea084 855 $chunk = $res[5];
856 $encoded = true;
08b7f7cc 857 }
858 if (!$encoded) {
bcae324b 859 if ($htmlsafe) {
9f7f68c3 860 $ret .= '&#32;';
08b7f7cc 861 } else {
862 $ret .= ' ';
da2415c1 863 }
08b7f7cc 864 }
dc3d13a7 865
bcae324b 866 if (!$encoded && $htmlsafe) {
dc3d13a7 867 $ret .= htmlspecialchars($chunk);
868 } else {
869 $ret .= $chunk;
870 }
098ea084 871 ++$i;
872 }
fd81e884 873 /* remove the first added space */
874 if ($ret) {
bcae324b 875 if ($htmlsafe) {
9f7f68c3 876 $ret = substr($ret,5);
fd81e884 877 } else {
878 $ret = substr($ret,1);
879 }
880 }
da2415c1 881
08b7f7cc 882 return $ret;
451f74a2 883}
884
9f7f68c3 885/**
a24cf710 886 * Encodes header
8bd0068d 887 *
a24cf710 888 * Function uses XTRA_CODE _encodeheader function, if such function exists.
a24cf710 889 *
758a7889 890 * Function uses Q encoding by default and encodes a string according to RFC
891 * 1522 for use in headers if it contains 8-bit characters or anything that
a24cf710 892 * looks like it should be encoded.
8bd0068d 893 *
758a7889 894 * Function switches to B encoding and encodeHeaderBase64() function, if
895 * string is 8bit and multibyte character set supported by mbstring extension
896 * is used. It can cause E_USER_NOTICE errors, if interface is used with
f270a6eb 897 * multibyte character set unsupported by mbstring extension.
898 *
8bd0068d 899 * @param string $string header string, that has to be encoded
900 * @return string quoted-printable encoded string
f270a6eb 901 * @todo make $mb_charsets system wide constant
8bd0068d 902 */
451f74a2 903function encodeHeader ($string) {
6fbd125b 904 global $default_charset, $languages, $squirrelmail_language;
83be314a 905
10dec454 906 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 907 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
33a55f5a 908 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
83be314a 909 }
793cc001 910
a24cf710 911 // Use B encoding for multibyte charsets
f270a6eb 912 $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
913 if (in_array($default_charset,$mb_charsets) &&
914 in_array($default_charset,sq_mb_list_encodings()) &&
915 sq_is8bit($string)) {
916 return encodeHeaderBase64($string,$default_charset);
917 } elseif (in_array($default_charset,$mb_charsets) &&
918 sq_is8bit($string) &&
919 ! in_array($default_charset,sq_mb_list_encodings())) {
920 // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
921 // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
922 }
a24cf710 923
451f74a2 924 // Encode only if the string contains 8-bit characters or =?
3d8371be 925 $j = strlen($string);
098ea084 926 $max_l = 75 - strlen($default_charset) - 7;
927 $aRet = array();
451f74a2 928 $ret = '';
c96c32f4 929 $iEncStart = $enc_init = false;
0d53f0f9 930 $cur_l = $iOffset = 0;
3d8371be 931 for($i = 0; $i < $j; ++$i) {
c96c32f4 932 switch($string{$i})
933 {
8bd0068d 934 case '=':
935 case '<':
936 case '>':
937 case ',':
938 case '?':
939 case '_':
940 if ($iEncStart === false) {
941 $iEncStart = $i;
942 }
943 $cur_l+=3;
944 if ($cur_l > ($max_l-2)) {
945 /* if there is an stringpart that doesn't need encoding, add it */
08b7f7cc 946 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
c96c32f4 947 $aRet[] = "=?$default_charset?Q?$ret?=";
948 $iOffset = $i;
949 $cur_l = 0;
950 $ret = '';
951 $iEncStart = false;
08b7f7cc 952 } else {
8bd0068d 953 $ret .= sprintf("=%02X",ord($string{$i}));
c96c32f4 954 }
8bd0068d 955 break;
956 case '(':
957 case ')':
958 if ($iEncStart !== false) {
08b7f7cc 959 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
8bd0068d 960 $aRet[] = "=?$default_charset?Q?$ret?=";
c96c32f4 961 $iOffset = $i;
8bd0068d 962 $cur_l = 0;
963 $ret = '';
964 $iEncStart = false;
c96c32f4 965 }
8bd0068d 966 break;
967 case ' ':
c96c32f4 968 if ($iEncStart !== false) {
098ea084 969 $cur_l++;
970 if ($cur_l > $max_l) {
08b7f7cc 971 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
c96c32f4 972 $aRet[] = "=?$default_charset?Q?$ret?=";
c96c32f4 973 $iOffset = $i;
974 $cur_l = 0;
098ea084 975 $ret = '';
8bd0068d 976 $iEncStart = false;
08b7f7cc 977 } else {
8bd0068d 978 $ret .= '_';
c96c32f4 979 }
3d8371be 980 }
8bd0068d 981 break;
982 default:
983 $k = ord($string{$i});
984 if ($k > 126) {
985 if ($iEncStart === false) {
986 // do not start encoding in the middle of a string, also take the rest of the word.
987 $sLeadString = substr($string,0,$i);
988 $aLeadString = explode(' ',$sLeadString);
989 $sToBeEncoded = array_pop($aLeadString);
990 $iEncStart = $i - strlen($sToBeEncoded);
991 $ret .= $sToBeEncoded;
992 $cur_l += strlen($sToBeEncoded);
993 }
994 $cur_l += 3;
995 /* first we add the encoded string that reached it's max size */
996 if ($cur_l > ($max_l-2)) {
997 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
998 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
999 $cur_l = 3;
1000 $ret = '';
1001 $iOffset = $i;
1002 $iEncStart = $i;
1003 }
1004 $enc_init = true;
1005 $ret .= sprintf("=%02X", $k);
1006 } else {
1007 if ($iEncStart !== false) {
1008 $cur_l++;
1009 if ($cur_l > $max_l) {
1010 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1011 $aRet[] = "=?$default_charset?Q?$ret?=";
1012 $iEncStart = false;
1013 $iOffset = $i;
1014 $cur_l = 0;
1015 $ret = '';
1016 } else {
1017 $ret .= $string{$i};
1018 }
1019 }
1020 }
1021 break;
f7b3ba37 1022 }
451f74a2 1023 }
793cc001 1024
c96c32f4 1025 if ($enc_init) {
1026 if ($iEncStart !== false) {
1027 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1028 $aRet[] = "=?$default_charset?Q?$ret?=";
1029 } else {
1030 $aRet[] = substr($string,$iOffset);
1031 }
1032 $string = implode('',$aRet);
451f74a2 1033 }
3d8371be 1034 return $string;
451f74a2 1035}
b74ba498 1036
f270a6eb 1037/**
1038 * Encodes string according to rfc2047 B encoding header formating rules
1039 *
758a7889 1040 * It is recommended way to encode headers with character sets that store
f270a6eb 1041 * symbols in more than one byte.
1042 *
1043 * Function requires mbstring support. If required mbstring functions are missing,
1044 * function returns false and sets E_USER_WARNING level error message.
1045 *
758a7889 1046 * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
1047 * that mbstring functions will generate E_WARNING errors, if unsupported
f270a6eb 1048 * character set is used. mb_encode_mimeheader function provided by php
1049 * mbstring extension is not used in order to get better control of header
1050 * encoding.
1051 *
758a7889 1052 * Used php code functions - function_exists(), trigger_error(), strlen()
1053 * (is used with charset names and base64 strings). Used php mbstring
f270a6eb 1054 * functions - mb_strlen and mb_substr.
1055 *
758a7889 1056 * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
f270a6eb 1057 * encoding), rfc 2822 (header folding)
1058 *
1059 * @param string $string header string that must be encoded
758a7889 1060 * @param string $charset character set. Must be supported by mbstring extension.
f270a6eb 1061 * Use sq_mb_list_encodings() to detect supported charsets.
1062 * @return string string encoded according to rfc2047 B encoding formating rules
1063 * @since 1.5.1
1064 * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1065 * @todo Do we want to control max length of header?
1066 * @todo Do we want to control EOL (end-of-line) marker?
1067 * @todo Do we want to translate error message?
1068 */
1069function encodeHeaderBase64($string,$charset) {
1070 /**
1071 * Check mbstring function requirements.
1072 */
1073 if (! function_exists('mb_strlen') ||
1074 ! function_exists('mb_substr')) {
1075 // set E_USER_WARNING
1076 trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
1077 // return false
1078 return false;
1079 }
1080
1081 // initial return array
1082 $aRet = array();
1083
1084 /**
1085 * header length = 75 symbols max (same as in encodeHeader)
1086 * remove $charset length
1087 * remove =? ? ?= (5 chars)
1088 * remove 2 more chars (\r\n ?)
1089 */
1090 $iMaxLength = 75 - strlen($charset) - 7;
1091
1092 // set first character position
1093 $iStartCharNum = 0;
1094
1095 // loop through all characters. count characters and not bytes.
1096 for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1097 // encode string from starting character to current character.
1098 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1099
1100 // Check encoded string length
1101 if(strlen($encoded_string)>$iMaxLength) {
1102 // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1103 $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1104
1105 // set new starting character
1106 $iStartCharNum = $iCharNum-1;
1107
1108 // encode last char (in case it is last character in string)
1109 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1110 } // if string is shorter than max length - add next character
1111 }
1112
1113 // add last encoded string to array
1114 $aRet[] = $encoded_string;
1115
1116 // set initial return string
1117 $sRet = '';
1118
1119 // loop through encoded strings
1120 foreach($aRet as $string) {
1121 // TODO: Do we want to control EOL (end-of-line) marker
1122 if ($sRet!='') $sRet.= " ";
1123
1124 // add header tags and encoded string to return string
1125 $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1126 }
1127
1128 return $sRet;
1129}
1130
691a2d25 1131/* This function trys to locate the entity_id of a specific mime element */
3d8371be 1132function find_ent_id($id, $message) {
a171b359 1133 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1134 if ($message->entities[$i]->header->type0 == 'multipart') {
3d8371be 1135 $ret = find_ent_id($id, $message->entities[$i]);
451f74a2 1136 } else {
3d8371be 1137 if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
d8cffbab 1138// if (sq_check_save_extension($message->entities[$i])) {
8bd0068d 1139 return $message->entities[$i]->entity_id;
da2415c1 1140// }
c8f5f606 1141 } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1142 /**
1143 * This is part of a fix for Outlook Express 6.x generating
1144 * cid URLs without creating content-id headers
1145 * @@JA - 20050207
1146 */
1147 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1148 return $message->entities[$i]->entity_id;
1149 }
3d8371be 1150 }
a3daaaf3 1151 }
451f74a2 1152 }
3d8371be 1153 return $ret;
451f74a2 1154}
a3daaaf3 1155
e5e9381a 1156function sq_check_save_extension($message) {
1157 $filename = $message->getFilename();
1158 $ext = substr($filename, strrpos($filename,'.')+1);
1159 $save_extensions = array('jpg','jpeg','gif','png','bmp');
3d8371be 1160 return in_array($ext, $save_extensions);
e5e9381a 1161}
1162
1163
691a2d25 1164/**
8bd0068d 1165 ** HTMLFILTER ROUTINES
1166 */
451f74a2 1167
2dd879b8 1168/**
0493ed11 1169 * This function checks attribute values for entity-encoded values
1170 * and returns them translated into 8-bit strings so we can run
1171 * checks on them.
8bd0068d 1172 *
0493ed11 1173 * @param $attvalue A string to run entity check against.
1174 * @return Nothing, modifies a reference value.
8bd0068d 1175 */
0493ed11 1176function sq_defang(&$attvalue){
1177 $me = 'sq_defang';
2dd879b8 1178 /**
0493ed11 1179 * Skip this if there aren't ampersands or backslashes.
8bd0068d 1180 */
0493ed11 1181 if (strpos($attvalue, '&') === false
1182 && strpos($attvalue, '\\') === false){
1183 return;
2dd879b8 1184 }
0493ed11 1185 $m = false;
2b665f28 1186 // before deent, translate the dangerous unicode characters and ... to safe values
1187 // otherwise the regular expressions do not match.
1188
1189
1190
0493ed11 1191 do {
1192 $m = false;
1193 $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1194 $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1195 $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1196 } while ($m == true);
1197 $attvalue = stripslashes($attvalue);
2dd879b8 1198}
1199
1200/**
8bd0068d 1201 * Kill any tabs, newlines, or carriage returns. Our friends the
1202 * makers of the browser with 95% market value decided that it'd
1203 * be funny to make "java[tab]script" be just as good as "javascript".
1204 *
1205 * @param attvalue The attribute value before extraneous spaces removed.
0493ed11 1206 * @return attvalue Nothing, modifies a reference value.
8bd0068d 1207 */
0493ed11 1208function sq_unspace(&$attvalue){
1209 $me = 'sq_unspace';
1210 if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1211 $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1212 Array('', '', '', '', ''), $attvalue);
2dd879b8 1213 }
2dd879b8 1214}
1215
2b665f28 1216/**
88de4926 1217 * Translate all dangerous Unicode or Shift_JIS characters which are accepted by
2b665f28 1218 * IE as regular characters.
1219 *
1220 * @param attvalue The attribute value before dangerous characters are translated.
1221 * @return attvalue Nothing, modifies a reference value.
1222 * @author Marc Groot Koerkamp.
1223 */
1224function sq_fixIE_idiocy(&$attvalue) {
1225 // remove NUL
1226 $attvalue = str_replace("\0", "", $attvalue);
1227 // remove comments
1228 $attvalue = preg_replace("/(\/\*.*?\*\/)/","",$attvalue);
1229
88de4926 1230 // IE has the evil habit of accepting every possible value for the attribute expression.
1231 // The table below contains characters which are parsed by IE if they are used in the "expression"
2b665f28 1232 // attribute value.
1233 $aDangerousCharsReplacementTable = array(
1234 array('&#x029F;', '&#0671;' ,/* L UNICODE IPA Extension */
1235 '&#x0280;', '&#0640;' ,/* R UNICODE IPA Extension */
1236 '&#x0274;', '&#0628;' ,/* N UNICODE IPA Extension */
567dc524 1237 '&#xFF25;', '&#65317;' ,/* Unicode FULLWIDTH LATIN CAPITAL LETTER E */
1238 '&#xFF45;', '&#65349;' ,/* Unicode FULLWIDTH LATIN SMALL LETTER E */
2b665f28 1239 '&#xFF38;', '&#65336;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER X */
1240 '&#xFF58;', '&#65368;',/* Unicode FULLWIDTH LATIN SMALL LETTER X */
1241 '&#xFF30;', '&#65328;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER P */
1242 '&#xFF50;', '&#65360;',/* Unicode FULLWIDTH LATIN SMALL LETTER P */
1243 '&#xFF32;', '&#65330;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER R */
1244 '&#xFF52;', '&#65362;',/* Unicode FULLWIDTH LATIN SMALL LETTER R */
1245 '&#xFF33;', '&#65331;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER S */
1246 '&#xFF53;', '&#65363;',/* Unicode FULLWIDTH LATIN SMALL LETTER S */
1247 '&#xFF29;', '&#65321;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER I */
1248 '&#xFF49;', '&#65353;',/* Unicode FULLWIDTH LATIN SMALL LETTER I */
1249 '&#xFF2F;', '&#65327;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER O */
1250 '&#xFF4F;', '&#65359;',/* Unicode FULLWIDTH LATIN SMALL LETTER O */
1251 '&#xFF2E;', '&#65326;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER N */
1252 '&#xFF4E;', '&#65358;',/* Unicode FULLWIDTH LATIN SMALL LETTER N */
1253 '&#xFF2C;', '&#65324;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER L */
1254 '&#xFF4C;', '&#65356;',/* Unicode FULLWIDTH LATIN SMALL LETTER L */
1255 '&#xFF35;', '&#65333;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER U */
1256 '&#xFF55;', '&#65365;',/* Unicode FULLWIDTH LATIN SMALL LETTER U */
1257 '&#x207F;', '&#8319;' ,/* Unicode SUPERSCRIPT LATIN SMALL LETTER N */
567dc524 1258 "\xEF\xBC\xA5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER E */ // in unicode this is some Chinese char range
1259 "\xEF\xBD\x85", /* Shift JIS FULLWIDTH LATIN SMALL LETTER E */
1260 "\xEF\xBC\xB8", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER X */
1261 "\xEF\xBD\x98", /* Shift JIS FULLWIDTH LATIN SMALL LETTER X */
1262 "\xEF\xBC\xB0", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER P */
1263 "\xEF\xBD\x90", /* Shift JIS FULLWIDTH LATIN SMALL LETTER P */
1264 "\xEF\xBC\xB2", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER R */
1265 "\xEF\xBD\x92", /* Shift JIS FULLWIDTH LATIN SMALL LETTER R */
1266 "\xEF\xBC\xB3", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER S */
1267 "\xEF\xBD\x93", /* Shift JIS FULLWIDTH LATIN SMALL LETTER S */
1268 "\xEF\xBC\xA9", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER I */
1269 "\xEF\xBD\x89", /* Shift JIS FULLWIDTH LATIN SMALL LETTER I */
1270 "\xEF\xBC\xAF", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER O */
1271 "\xEF\xBD\x8F", /* Shift JIS FULLWIDTH LATIN SMALL LETTER O */
1272 "\xEF\xBC\xAE", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER N */
1273 "\xEF\xBD\x8E", /* Shift JIS FULLWIDTH LATIN SMALL LETTER N */
1274 "\xEF\xBC\xAC", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER L */
1275 "\xEF\xBD\x8C", /* Shift JIS FULLWIDTH LATIN SMALL LETTER L */
1276 "\xEF\xBC\xB5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER U */
1277 "\xEF\xBD\x95", /* Shift JIS FULLWIDTH LATIN SMALL LETTER U */
1278 "\xE2\x81\xBF", /* Shift JIS FULLWIDTH SUPERSCRIPT N */
1279 "\xCA\x9F", /* L UNICODE IPA Extension */
1280 "\xCA\x80", /* R UNICODE IPA Extension */
1281 "\xC9\xB4"), /* N UNICODE IPA Extension */
2b665f28 1282 array('l', 'l', 'r','r','n','n',
567dc524 1283 'E','E','e','e','X','X','x','x','P','P','p','p','R','R','r','r','S','S','s','s','I','I',
1284 'i','i','O','O','o','o','N','N','n','n','L','L','l','l','U','U','u','u','n','n',
1285 '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 1286 $attvalue = str_replace($aDangerousCharsReplacementTable[0],$aDangerousCharsReplacementTable[1],$attvalue);
1287
88de4926 1288 // Escapes are useful for special characters like "{}[]()'&. In other cases they are
1289 // used for XSS.
2b665f28 1290 $attvalue = preg_replace("/(\\\\)([a-zA-Z]{1})/",'$2',$attvalue);
1291}
1292
691a2d25 1293/**
8bd0068d 1294 * This function returns the final tag out of the tag name, an array
1295 * of attributes, and the type of the tag. This function is called by
1296 * sq_sanitize internally.
1297 *
1298 * @param $tagname the name of the tag.
1299 * @param $attary the array of attributes and their values
1300 * @param $tagtype The type of the tag (see in comments).
1301 * @return a string with the final tag representation.
1302 */
691a2d25 1303function sq_tagprint($tagname, $attary, $tagtype){
b583c3e8 1304 $me = 'sq_tagprint';
3d8371be 1305
691a2d25 1306 if ($tagtype == 2){
1307 $fulltag = '</' . $tagname . '>';
1308 } else {
1309 $fulltag = '<' . $tagname;
1310 if (is_array($attary) && sizeof($attary)){
1311 $atts = Array();
1312 while (list($attname, $attvalue) = each($attary)){
1313 array_push($atts, "$attname=$attvalue");
1314 }
1315 $fulltag .= ' ' . join(" ", $atts);
1316 }
1317 if ($tagtype == 3){
b583c3e8 1318 $fulltag .= ' /';
691a2d25 1319 }
b583c3e8 1320 $fulltag .= '>';
451f74a2 1321 }
691a2d25 1322 return $fulltag;
451f74a2 1323}
a3daaaf3 1324
691a2d25 1325/**
8bd0068d 1326 * A small helper function to use with array_walk. Modifies a by-ref
1327 * value and makes it lowercase.
1328 *
1329 * @param $val a value passed by-ref.
1330 * @return void since it modifies a by-ref value.
1331 */
691a2d25 1332function sq_casenormalize(&$val){
1333 $val = strtolower($val);
1334}
451f74a2 1335
691a2d25 1336/**
8bd0068d 1337 * This function skips any whitespace from the current position within
1338 * a string and to the next non-whitespace value.
1339 *
1340 * @param $body the string
1341 * @param $offset the offset within the string where we should start
1342 * looking for the next non-whitespace character.
1343 * @return the location within the $body where the next
1344 * non-whitespace char is located.
1345 */
691a2d25 1346function sq_skipspace($body, $offset){
b583c3e8 1347 $me = 'sq_skipspace';
3d8371be 1348 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
691a2d25 1349 if (sizeof($matches{1})){
1350 $count = strlen($matches{1});
1351 $offset += $count;
451f74a2 1352 }
691a2d25 1353 return $offset;
451f74a2 1354}
a3daaaf3 1355
691a2d25 1356/**
8bd0068d 1357 * This function looks for the next character within a string. It's
1358 * really just a glorified "strpos", except it catches if failures
1359 * nicely.
1360 *
1361 * @param $body The string to look for needle in.
1362 * @param $offset Start looking from this position.
1363 * @param $needle The character/string to look for.
1364 * @return location of the next occurance of the needle, or
1365 * strlen($body) if needle wasn't found.
1366 */
691a2d25 1367function sq_findnxstr($body, $offset, $needle){
3d8371be 1368 $me = 'sq_findnxstr';
691a2d25 1369 $pos = strpos($body, $needle, $offset);
1370 if ($pos === FALSE){
1371 $pos = strlen($body);
451f74a2 1372 }
691a2d25 1373 return $pos;
451f74a2 1374}
a3daaaf3 1375
691a2d25 1376/**
8bd0068d 1377 * This function takes a PCRE-style regexp and tries to match it
1378 * within the string.
1379 *
1380 * @param $body The string to look for needle in.
1381 * @param $offset Start looking from here.
1382 * @param $reg A PCRE-style regex to match.
1383 * @return Returns a false if no matches found, or an array
1384 * with the following members:
1385 * - integer with the location of the match within $body
1386 * - string with whatever content between offset and the match
1387 * - string with whatever it is we matched
1388 */
691a2d25 1389function sq_findnxreg($body, $offset, $reg){
b583c3e8 1390 $me = 'sq_findnxreg';
691a2d25 1391 $matches = Array();
1392 $retarr = Array();
7d06541f 1393 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1394 if (!isset($matches{0}) || !$matches{0}){
691a2d25 1395 $retarr = false;
1396 } else {
1397 $retarr{0} = $offset + strlen($matches{1});
1398 $retarr{1} = $matches{1};
1399 $retarr{2} = $matches{2};
1400 }
1401 return $retarr;
1402}
a3daaaf3 1403
691a2d25 1404/**
8bd0068d 1405 * This function looks for the next tag.
1406 *
1407 * @param $body String where to look for the next tag.
1408 * @param $offset Start looking from here.
1409 * @return false if no more tags exist in the body, or
1410 * an array with the following members:
1411 * - string with the name of the tag
1412 * - array with attributes and their values
1413 * - integer with tag type (1, 2, or 3)
1414 * - integer where the tag starts (starting "<")
1415 * - integer where the tag ends (ending ">")
1416 * first three members will be false, if the tag is invalid.
1417 */
691a2d25 1418function sq_getnxtag($body, $offset){
b583c3e8 1419 $me = 'sq_getnxtag';
691a2d25 1420 if ($offset > strlen($body)){
1421 return false;
1422 }
1423 $lt = sq_findnxstr($body, $offset, "<");
1424 if ($lt == strlen($body)){
1425 return false;
1426 }
1427 /**
8bd0068d 1428 * We are here:
1429 * blah blah <tag attribute="value">
1430 * \---------^
1431 */
691a2d25 1432 $pos = sq_skipspace($body, $lt+1);
1433 if ($pos >= strlen($body)){
1434 return Array(false, false, false, $lt, strlen($body));
1435 }
1436 /**
8bd0068d 1437 * There are 3 kinds of tags:
1438 * 1. Opening tag, e.g.:
1439 * <a href="blah">
1440 * 2. Closing tag, e.g.:
1441 * </a>
1442 * 3. XHTML-style content-less tag, e.g.:
1443 * <img src="blah" />
1444 */
691a2d25 1445 $tagtype = false;
1446 switch (substr($body, $pos, 1)){
3d8371be 1447 case '/':
1448 $tagtype = 2;
1449 $pos++;
1450 break;
1451 case '!':
1452 /**
8bd0068d 1453 * A comment or an SGML declaration.
1454 */
3d8371be 1455 if (substr($body, $pos+1, 2) == "--"){
1456 $gt = strpos($body, "-->", $pos);
1457 if ($gt === false){
1458 $gt = strlen($body);
1459 } else {
1460 $gt += 2;
1461 }
1462 return Array(false, false, false, $lt, $gt);
bb8d0799 1463 } else {
3d8371be 1464 $gt = sq_findnxstr($body, $pos, ">");
1465 return Array(false, false, false, $lt, $gt);
1466 }
1467 break;
1468 default:
1469 /**
8bd0068d 1470 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1471 * later.
1472 */
3d8371be 1473 $tagtype = 1;
1474 break;
691a2d25 1475 }
a3daaaf3 1476
0493ed11 1477 $tag_start = $pos;
691a2d25 1478 $tagname = '';
1479 /**
8bd0068d 1480 * Look for next [\W-_], which will indicate the end of the tag name.
1481 */
691a2d25 1482 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1483 if ($regary == false){
1484 return Array(false, false, false, $lt, strlen($body));
1485 }
1486 list($pos, $tagname, $match) = $regary;
1487 $tagname = strtolower($tagname);
1488
1489 /**
8bd0068d 1490 * $match can be either of these:
1491 * '>' indicating the end of the tag entirely.
1492 * '\s' indicating the end of the tag name.
1493 * '/' indicating that this is type-3 xhtml tag.
1494 *
1495 * Whatever else we find there indicates an invalid tag.
1496 */
691a2d25 1497 switch ($match){
3d8371be 1498 case '/':
691a2d25 1499 /**
8bd0068d 1500 * This is an xhtml-style tag with a closing / at the
1501 * end, like so: <img src="blah" />. Check if it's followed
1502 * by the closing bracket. If not, then this tag is invalid
1503 */
3d8371be 1504 if (substr($body, $pos, 2) == "/>"){
1505 $pos++;
1506 $tagtype = 3;
1507 } else {
1508 $gt = sq_findnxstr($body, $pos, ">");
1509 $retary = Array(false, false, false, $lt, $gt);
1510 return $retary;
1511 }
1512 case '>':
1513 return Array($tagname, false, $tagtype, $lt, $pos);
1514 break;
1515 default:
1516 /**
8bd0068d 1517 * Check if it's whitespace
1518 */
3d8371be 1519 if (!preg_match('/\s/', $match)){
1520 /**
8bd0068d 1521 * This is an invalid tag! Look for the next closing ">".
1522 */
7d06541f 1523 $gt = sq_findnxstr($body, $lt, ">");
3d8371be 1524 return Array(false, false, false, $lt, $gt);
1525 }
1526 break;
691a2d25 1527 }
3d8371be 1528
691a2d25 1529 /**
8bd0068d 1530 * At this point we're here:
1531 * <tagname attribute='blah'>
1532 * \-------^
1533 *
1534 * At this point we loop in order to find all attributes.
1535 */
691a2d25 1536 $attname = '';
0493ed11 1537 $atttype = false;
691a2d25 1538 $attary = Array();
1539
1540 while ($pos <= strlen($body)){
1541 $pos = sq_skipspace($body, $pos);
1542 if ($pos == strlen($body)){
1543 /**
8bd0068d 1544 * Non-closed tag.
1545 */
691a2d25 1546 return Array(false, false, false, $lt, $pos);
1547 }
1548 /**
8bd0068d 1549 * See if we arrived at a ">" or "/>", which means that we reached
1550 * the end of the tag.
1551 */
691a2d25 1552 $matches = Array();
164800ad 1553 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
c828931c 1554 /**
8bd0068d 1555 * Yep. So we did.
1556 */
c828931c 1557 $pos += strlen($matches{1});
1558 if ($matches{2} == "/>"){
1559 $tagtype = 3;
1560 $pos++;
1561 }
1562 return Array($tagname, $attary, $tagtype, $lt, $pos);
1563 }
a3daaaf3 1564
cca46357 1565 /**
8bd0068d 1566 * There are several types of attributes, with optional
1567 * [:space:] between members.
1568 * Type 1:
1569 * attrname[:space:]=[:space:]'CDATA'
1570 * Type 2:
1571 * attrname[:space:]=[:space:]"CDATA"
1572 * Type 3:
1573 * attr[:space:]=[:space:]CDATA
1574 * Type 4:
1575 * attrname
1576 *
1577 * We leave types 1 and 2 the same, type 3 we check for
1578 * '"' and convert to "&quot" if needed, then wrap in
1579 * double quotes. Type 4 we convert into:
1580 * attrname="yes".
1581 */
3f7c623f 1582 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
691a2d25 1583 if ($regary == false){
1584 /**
8bd0068d 1585 * Looks like body ended before the end of tag.
1586 */
691a2d25 1587 return Array(false, false, false, $lt, strlen($body));
cca46357 1588 }
691a2d25 1589 list($pos, $attname, $match) = $regary;
1590 $attname = strtolower($attname);
1591 /**
8bd0068d 1592 * We arrived at the end of attribute name. Several things possible
1593 * here:
1594 * '>' means the end of the tag and this is attribute type 4
1595 * '/' if followed by '>' means the same thing as above
1596 * '\s' means a lot of things -- look what it's followed by.
1597 * anything else means the attribute is invalid.
1598 */
691a2d25 1599 switch($match){
3d8371be 1600 case '/':
691a2d25 1601 /**
8bd0068d 1602 * This is an xhtml-style tag with a closing / at the
1603 * end, like so: <img src="blah" />. Check if it's followed
1604 * by the closing bracket. If not, then this tag is invalid
1605 */
3d8371be 1606 if (substr($body, $pos, 2) == "/>"){
691a2d25 1607 $pos++;
3d8371be 1608 $tagtype = 3;
691a2d25 1609 } else {
3d8371be 1610 $gt = sq_findnxstr($body, $pos, ">");
1611 $retary = Array(false, false, false, $lt, $gt);
1612 return $retary;
1613 }
1614 case '>':
1615 $attary{$attname} = '"yes"';
1616 return Array($tagname, $attary, $tagtype, $lt, $pos);
1617 break;
1618 default:
1619 /**
8bd0068d 1620 * Skip whitespace and see what we arrive at.
1621 */
3d8371be 1622 $pos = sq_skipspace($body, $pos);
1623 $char = substr($body, $pos, 1);
1624 /**
8bd0068d 1625 * Two things are valid here:
1626 * '=' means this is attribute type 1 2 or 3.
1627 * \w means this was attribute type 4.
1628 * anything else we ignore and re-loop. End of tag and
1629 * invalid stuff will be caught by our checks at the beginning
1630 * of the loop.
1631 */
3d8371be 1632 if ($char == "="){
1633 $pos++;
1634 $pos = sq_skipspace($body, $pos);
691a2d25 1635 /**
8bd0068d 1636 * Here are 3 possibilities:
1637 * "'" attribute type 1
1638 * '"' attribute type 2
1639 * everything else is the content of tag type 3
1640 */
3d8371be 1641 $quot = substr($body, $pos, 1);
1642 if ($quot == "'"){
1643 $regary = sq_findnxreg($body, $pos+1, "\'");
1644 if ($regary == false){
1645 return Array(false, false, false, $lt, strlen($body));
1646 }
1647 list($pos, $attval, $match) = $regary;
1648 $pos++;
1649 $attary{$attname} = "'" . $attval . "'";
1650 } else if ($quot == '"'){
1651 $regary = sq_findnxreg($body, $pos+1, '\"');
1652 if ($regary == false){
1653 return Array(false, false, false, $lt, strlen($body));
1654 }
1655 list($pos, $attval, $match) = $regary;
1656 $pos++;
1657 $attary{$attname} = '"' . $attval . '"';
1658 } else {
1659 /**
8bd0068d 1660 * These are hateful. Look for \s, or >.
1661 */
3d8371be 1662 $regary = sq_findnxreg($body, $pos, "[\s>]");
1663 if ($regary == false){
1664 return Array(false, false, false, $lt, strlen($body));
1665 }
1666 list($pos, $attval, $match) = $regary;
1667 /**
8bd0068d 1668 * If it's ">" it will be caught at the top.
1669 */
3d8371be 1670 $attval = preg_replace("/\"/s", "&quot;", $attval);
1671 $attary{$attname} = '"' . $attval . '"';
7e235a1a 1672 }
3d8371be 1673 } else if (preg_match("|[\w/>]|", $char)) {
691a2d25 1674 /**
8bd0068d 1675 * That was attribute type 4.
1676 */
3d8371be 1677 $attary{$attname} = '"yes"';
1678 } else {
1679 /**
8bd0068d 1680 * An illegal character. Find next '>' and return.
1681 */
3d8371be 1682 $gt = sq_findnxstr($body, $pos, ">");
1683 return Array(false, false, false, $lt, $gt);
451f74a2 1684 }
3d8371be 1685 break;
691a2d25 1686 }
1687 }
1688 /**
8bd0068d 1689 * The fact that we got here indicates that the tag end was never
1690 * found. Return invalid tag indication so it gets stripped.
1691 */
691a2d25 1692 return Array(false, false, false, $lt, strlen($body));
1693}
1694
1695/**
0493ed11 1696 * Translates entities into literal values so they can be checked.
8bd0068d 1697 *
0493ed11 1698 * @param $attvalue the by-ref value to check.
1699 * @param $regex the regular expression to check against.
1700 * @param $hex whether the entites are hexadecimal.
1701 * @return True or False depending on whether there were matches.
8bd0068d 1702 */
0493ed11 1703function sq_deent(&$attvalue, $regex, $hex=false){
b583c3e8 1704 $me = 'sq_deent';
0493ed11 1705 $ret_match = false;
2b665f28 1706 // remove comments
1707 //$attvalue = preg_replace("/(\/\*.*\*\/)/","",$attvalue);
0493ed11 1708 preg_match_all($regex, $attvalue, $matches);
1709 if (is_array($matches) && sizeof($matches[0]) > 0){
1710 $repl = Array();
1711 for ($i = 0; $i < sizeof($matches[0]); $i++){
1712 $numval = $matches[1][$i];
1713 if ($hex){
1714 $numval = hexdec($numval);
a3daaaf3 1715 }
0493ed11 1716 $repl{$matches[0][$i]} = chr($numval);
691a2d25 1717 }
0493ed11 1718 $attvalue = strtr($attvalue, $repl);
1719 return true;
1720 } else {
1721 return false;
691a2d25 1722 }
691a2d25 1723}
1724
1725/**
8bd0068d 1726 * This function runs various checks against the attributes.
1727 *
1728 * @param $tagname String with the name of the tag.
1729 * @param $attary Array with all tag attributes.
1730 * @param $rm_attnames See description for sq_sanitize
1731 * @param $bad_attvals See description for sq_sanitize
1732 * @param $add_attr_to_tag See description for sq_sanitize
1733 * @param $message message object
1734 * @param $id message id
1735 * @return Array with modified attributes.
1736 */
da2415c1 1737function sq_fixatts($tagname,
1738 $attary,
691a2d25 1739 $rm_attnames,
1740 $bad_attvals,
1741 $add_attr_to_tag,
1742 $message,
b3af12ef 1743 $id,
3d8371be 1744 $mailbox
691a2d25 1745 ){
b583c3e8 1746 $me = 'sq_fixatts';
691a2d25 1747 while (list($attname, $attvalue) = each($attary)){
1748 /**
8bd0068d 1749 * See if this attribute should be removed.
1750 */
691a2d25 1751 foreach ($rm_attnames as $matchtag=>$matchattrs){
1752 if (preg_match($matchtag, $tagname)){
1753 foreach ($matchattrs as $matchattr){
1754 if (preg_match($matchattr, $attname)){
1755 unset($attary{$attname});
1756 continue;
1757 }
451f74a2 1758 }
451f74a2 1759 }
691a2d25 1760 }
2b665f28 1761 /**
1762 * Workaround for IE quirks
1763 */
1764 sq_fixIE_idiocy($attvalue);
1765
691a2d25 1766 /**
8bd0068d 1767 * Remove any backslashes, entities, and extraneous whitespace.
1768 */
2b665f28 1769
1770 $oldattvalue = $attvalue;
0493ed11 1771 sq_defang($attvalue);
2b665f28 1772 if ($attname == 'style' && $attvalue !== $oldattvalue) {
1773 // entities are used in the attribute value. In 99% of the cases it's there as XSS
1774 // i.e.<div style="{ left:exp&#x0280;essio&#x0274;( alert('XSS') ) }">
1775 $attvalue = "idiocy";
1776 $attary{$attname} = $attvalue;
1777 }
0493ed11 1778 sq_unspace($attvalue);
af861a34 1779
691a2d25 1780 /**
8bd0068d 1781 * Now let's run checks on the attvalues.
1782 * I don't expect anyone to comprehend this. If you do,
1783 * get in touch with me so I can drive to where you live and
1784 * shake your hand personally. :)
1785 */
691a2d25 1786 foreach ($bad_attvals as $matchtag=>$matchattrs){
1787 if (preg_match($matchtag, $tagname)){
1788 foreach ($matchattrs as $matchattr=>$valary){
1789 if (preg_match($matchattr, $attname)){
1790 /**
8bd0068d 1791 * There are two arrays in valary.
1792 * First is matches.
1793 * Second one is replacements
1794 */
691a2d25 1795 list($valmatch, $valrepl) = $valary;
da2415c1 1796 $newvalue =
691a2d25 1797 preg_replace($valmatch, $valrepl, $attvalue);
1798 if ($newvalue != $attvalue){
1799 $attary{$attname} = $newvalue;
567dc524 1800 $attvalue = $newvalue;
691a2d25 1801 }
1802 }
1803 }
451f74a2 1804 }
a3daaaf3 1805 }
567dc524 1806 if ($attname == 'style') {
1807 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1808 // 8bit and control characters in style attribute values can be used for XSS, remove them
1809 $attary{$attname} = '"disallowed character"';
1810 }
1811 preg_match_all("/url\s*\((.+)\)/si",$attvalue,$aMatch);
1812 if (count($aMatch)) {
1813 foreach($aMatch[1] as $sMatch) {
1814 // url value
1815 $urlvalue = $sMatch;
1816 sq_fix_url($attname, $urlvalue, $message, $id, $mailbox,"'");
1817 $attary{$attname} = str_replace($sMatch,$urlvalue,$attvalue);
1818 }
1819 }
691a2d25 1820 }
ff940ebc 1821 /**
567dc524 1822 * Use white list based filtering on attributes which can contain url's
ff940ebc 1823 */
567dc524 1824 else if ($attname == 'href' || $attname == 'src' || $attname == 'background') {
1825 sq_fix_url($attname, $attvalue, $message, $id, $mailbox);
1826 $attary{$attname} = $attvalue;
ff940ebc 1827 }
a3daaaf3 1828 }
691a2d25 1829 /**
8bd0068d 1830 * See if we need to append any attributes to this tag.
1831 */
691a2d25 1832 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1833 if (preg_match($matchtag, $tagname)){
1834 $attary = array_merge($attary, $addattary);
1835 }
1836 }
1837 return $attary;
451f74a2 1838}
a3daaaf3 1839
567dc524 1840/**
1841 * This function filters url's
1842 *
1843 * @param $attvalue String with attribute value to filter
1844 * @param $message message object
1845 * @param $id message id
1846 * @param $mailbox mailbox
1847 * @param $sQuote quoting characters around url's
1848 */
1849function sq_fix_url($attname, &$attvalue, $message, $id, $mailbox,$sQuote = '"') {
1850 $attvalue = trim($attvalue);
1851 if ($attvalue && ($attvalue[0] =='"'|| $attvalue[0] == "'")) {
1852 // remove the double quotes
1853 $sQuote = $attvalue[0];
1854 $attvalue = trim(substr($attvalue,1,-1));
1855 }
1856
9ebace19 1857 // If there's no "view_unsafe_images" setting in the user's preferences,
1858 // turn unsafe images off by default.
567dc524 1859 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1860 $view_unsafe_images = false;
1861 }
1862 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1863
1864 /**
1865 * Replace empty src tags with the blank image. src is only used
1866 * for frames, images, and image inputs. Doing a replace should
1867 * not affect them working as should be, however it will stop
1868 * IE from being kicked off when src for img tags are not set
1869 */
1870 if ($attvalue == '') {
1871 $attvalue = '"' . SM_PATH . 'images/blank.png"';
1872 } else {
1873 // first, disallow 8 bit characters and control characters
1874 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1875 switch ($attname) {
1876 case 'href':
1877 $attvalue = $sQuote . 'http://invalid-stuff-detected.example.com' . $sQuote;
1878 break;
1879 default:
1880 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1881 break;
1882 }
1883 } else {
1884 $aUrl = parse_url($attvalue);
1885 if (isset($aUrl['scheme'])) {
1886 switch(strtolower($aUrl['scheme'])) {
d75e755b 1887 case 'mailto':
567dc524 1888 case 'http':
1889 case 'https':
1890 case 'ftp':
1891 if ($attname != 'href') {
1892 if ($view_unsafe_images == false) {
1893 $attvalue = $sQuote . $secremoveimg . $sQuote;
1894 } else {
1895 if (isset($aUrl['path'])) {
1896 // validate image extension.
1897 $ext = strtolower(substr($aUrl['path'],strrpos($aUrl['path'],'.')));
1898 if (!in_array($ext,array('.jpeg','.jpg','xjpeg','.gif','.bmp','.jpe','.png','.xbm'))) {
1899 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1900 }
1901 } else {
1902 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1903 }
1904 }
1905 }
1906 break;
1907 case 'outbind':
1908 /**
1909 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1910 * One day MS might actually make it match something useful, for now, falling
1911 * back to using cid2http, so we can grab the blank.png.
1912 */
1913 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1914 break;
1915 case 'cid':
1916 /**
1917 * Turn cid: urls into http-friendly ones.
1918 */
1919 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1920 break;
1921 default:
1922 $attvalue = $sQuote . SM_PATH . 'images/blank.png' . $sQuote;
1923 break;
1924 }
1925 } else {
1926 if (!(isset($aUrl['path']) && $aUrl['path'] == $secremoveimg)) {
1927 // parse_url did not lead to satisfying result
1928 $attvalue = $sQuote . SM_PATH . 'images/blank.png' . $sQuote;
1929 }
1930 }
1931 }
1932 }
1933}
1934
691a2d25 1935/**
8bd0068d 1936 * This function edits the style definition to make them friendly and
1937 * usable in SquirrelMail.
1938 *
1939 * @param $message the message object
1940 * @param $id the message id
1941 * @param $content a string with whatever is between <style> and </style>
1942 * @param $mailbox the message mailbox
1943 * @return a string with edited content.
1944 */
e60a299a 1945function sq_fixstyle($body, $pos, $message, $id, $mailbox){
b583c3e8 1946 $me = 'sq_fixstyle';
7e2ff844 1947 // workaround for </style> in between comments
1948 $iCurrentPos = $pos;
1949 $content = '';
1950 $sToken = '';
1951 $bSucces = false;
1952 $bEndTag = false;
1953 for ($i=$pos,$iCount=strlen($body);$i<$iCount;++$i) {
1954 $char = $body{$i};
1955 switch ($char) {
1956 case '<':
10c0caeb 1957 $sToken = $char;
7e2ff844 1958 break;
1959 case '/':
1960 if ($sToken == '<') {
1961 $sToken .= $char;
1962 $bEndTag = true;
1963 } else {
1964 $content .= $char;
1965 }
1966 break;
1967 case '>':
1968 if ($bEndTag) {
1969 $sToken .= $char;
1970 if (preg_match('/\<\/\s*style\s*\>/i',$sToken,$aMatch)) {
1971 $newpos = $i + 1;
1972 $bSucces = true;
1973 break 2;
1974 } else {
1975 $content .= $sToken;
1976 }
1977 $bEndTag = false;
1978 } else {
1979 $content .= $char;
1980 }
1981 break;
1982 case '!':
1983 if ($sToken == '<') {
1984 // possible comment
1985 if (isset($body{$i+2}) && substr($body,$i,3) == '!--') {
1986 $i = strpos($body,'-->',$i+3);
2b665f28 1987 if ($i === false) { // no end comment
1988 $i = strlen($body);
1989 }
7e2ff844 1990 $sToken = '';
1991 }
1992 } else {
1993 $content .= $char;
1994 }
1995 break;
1996 default:
1997 if ($bEndTag) {
1998 $sToken .= $char;
1999 } else {
2000 $content .= $char;
2001 }
2002 break;
2003 }
2004 }
2005 if ($bSucces == FALSE){
7d06541f 2006 return array(FALSE, strlen($body));
2007 }
7e2ff844 2008
2b665f28 2009
2010
691a2d25 2011 /**
9ebace19 2012 * First look for general BODY style declaration, which would be
2013 * like so:
2014 * body {background: blah-blah}
2015 * and change it to .bodyclass so we can just assign it to a <div>
2016 */
691a2d25 2017 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
3d8371be 2018 $secremoveimg = '../images/' . _("sec_remove_eng.png");
691a2d25 2019 /**
0493ed11 2020 * Fix url('blah') declarations.
2021 */
2022 // $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
2023 // "url(\\1$secremoveimg\\2)", $content);
2b665f28 2024
567dc524 2025 // first check for 8bit sequences and disallowed control characters
2026 if (preg_match('/[\16-\37\200-\377]+/',$content)) {
2027 $content = '<!-- style block removed by html filter due to presence of 8bit characters -->';
2028 return array($content, $newpos);
2029 }
2030
2b665f28 2031 // IE Sucks hard. We have a special function for it.
2032 sq_fixIE_idiocy($content);
2033
2034 // remove @import line
2035 $content = preg_replace("/^\s*(@import.*)$/mi","\n<!-- @import rules forbidden -->\n",$content);
2036
5b4884be 2037 // translate ur\l and variations (IE parses that)
2b665f28 2038 // TODO check if the sq_fixIE_idiocy function already handles this.
5b4884be 2039 $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
567dc524 2040 preg_match_all("/url\s*\((.+)\)/si",$content,$aMatch);
2041 if (count($aMatch)) {
2042 $aValue = $aReplace = array();
2043 foreach($aMatch[1] as $sMatch) {
2044 // url value
2045 $urlvalue = $sMatch;
2046 sq_fix_url('style',$urlvalue, $message, $id, $mailbox,"'");
2047 $aValue[] = $sMatch;
2048 $aReplace[] = $urlvalue;
0493ed11 2049 }
567dc524 2050 $content = str_replace($aValue,$aReplace,$content);
691a2d25 2051 }
567dc524 2052
9ebace19 2053 /**
2054 * Remove any backslashes, entities, and extraneous whitespace.
2055 */
0493ed11 2056 $contentTemp = $content;
2057 sq_defang($contentTemp);
2058 sq_unspace($contentTemp);
a3daaaf3 2059
691a2d25 2060 /**
8bd0068d 2061 * Fix stupid css declarations which lead to vulnerabilities
2062 * in IE.
2063 */
5db90261 2064 $match = Array('/\/\*.*\*\//',
2065 '/expression/i',
0493ed11 2066 '/behaviou*r/i',
2067 '/binding/i',
2b665f28 2068 '/include-source/i',
2069 '/javascript/i',
2070 '/script/i');
2071 $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy');
0493ed11 2072 $contentNew = preg_replace($match, $replace, $contentTemp);
2073 if ($contentNew !== $contentTemp) {
2074 // insecure css declarations are used. From now on we don't care
2075 // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
2076 $content = $contentNew;
2077 }
7d06541f 2078 return array($content, $newpos);
691a2d25 2079}
a3daaaf3 2080
0493ed11 2081
691a2d25 2082/**
8bd0068d 2083 * This function converts cid: url's into the ones that can be viewed in
2084 * the browser.
2085 *
2086 * @param $message the message object
2087 * @param $id the message id
2088 * @param $cidurl the cid: url.
2089 * @param $mailbox the message mailbox
2090 * @return a string with a http-friendly url
2091 */
b3af12ef 2092function sq_cid2http($message, $id, $cidurl, $mailbox){
691a2d25 2093 /**
8bd0068d 2094 * Get rid of quotes.
2095 */
691a2d25 2096 $quotchar = substr($cidurl, 0, 1);
2dd879b8 2097 if ($quotchar == '"' || $quotchar == "'"){
2098 $cidurl = str_replace($quotchar, "", $cidurl);
2099 } else {
2100 $quotchar = '';
2101 }
691a2d25 2102 $cidurl = substr(trim($cidurl), 4);
0493ed11 2103
2104 $match_str = '/\{.*?\}\//';
2105 $str_rep = '';
2106 $cidurl = preg_replace($match_str, $str_rep, $cidurl);
2107
e5e9381a 2108 $linkurl = find_ent_id($cidurl, $message);
9ebace19 2109 /* in case of non-safe cid links $httpurl should be replaced by a sort of
2110 unsafe link image */
e5e9381a 2111 $httpurl = '';
c8f5f606 2112
8bd0068d 2113 /**
2114 * This is part of a fix for Outlook Express 6.x generating
2115 * cid URLs without creating content-id headers. These images are
2116 * not part of the multipart/related html mail. The html contains
2117 * <img src="cid:{some_id}/image_filename.ext"> references to
2118 * attached images with as goal to render them inline although
2119 * the attachment disposition property is not inline.
2120 */
c8f5f606 2121
2122 if (empty($linkurl)) {
2123 if (preg_match('/{.*}\//', $cidurl)) {
2124 $cidurl = preg_replace('/{.*}\//','', $cidurl);
2125 if (!empty($cidurl)) {
2126 $linkurl = find_ent_id($cidurl, $message);
2127 }
2128 }
2129 }
f8a1ed5a 2130
c8f5f606 2131 if (!empty($linkurl)) {
6b04287c 2132 $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
8bd0068d 2133 "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
2134 '&amp;ent_id=' . $linkurl . $quotchar;
c8f5f606 2135 } else {
2136 /**
2137 * If we couldn't generate a proper img url, drop in a blank image
2138 * instead of sending back empty, otherwise it causes unusual behaviour
2139 */
bc017c1d 2140 $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
e5e9381a 2141 }
f8a1ed5a 2142
691a2d25 2143 return $httpurl;
2144}
2145
2146/**
8bd0068d 2147 * This function changes the <body> tag into a <div> tag since we
2148 * can't really have a body-within-body.
2149 *
2150 * @param $attary an array of attributes and values of <body>
2151 * @param $mailbox mailbox we're currently reading (for cid2http)
2152 * @param $message current message (for cid2http)
2153 * @param $id current message id (for cid2http)
2154 * @return a modified array of attributes to be set for <div>
2155 */
2dd879b8 2156function sq_body2div($attary, $mailbox, $message, $id){
b583c3e8 2157 $me = 'sq_body2div';
3d8371be 2158 $divattary = Array('class' => "'bodyclass'");
b583c3e8 2159 $text = '#000000';
c189a963 2160 $has_bgc_stl = $has_txt_stl = false;
b583c3e8 2161 $styledef = '';
691a2d25 2162 if (is_array($attary) && sizeof($attary) > 0){
2163 foreach ($attary as $attname=>$attvalue){
2164 $quotchar = substr($attvalue, 0, 1);
2165 $attvalue = str_replace($quotchar, "", $attvalue);
2166 switch ($attname){
3d8371be 2167 case 'background':
8bd0068d 2168 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
3d8371be 2169 $styledef .= "background-image: url('$attvalue'); ";
2170 break;
2171 case 'bgcolor':
c189a963 2172 $has_bgc_stl = true;
3d8371be 2173 $styledef .= "background-color: $attvalue; ";
2174 break;
2175 case 'text':
c189a963 2176 $has_txt_stl = true;
3d8371be 2177 $styledef .= "color: $attvalue; ";
2178 break;
691a2d25 2179 }
a3daaaf3 2180 }
c189a963 2181 // Outlook defines a white bgcolor and no text color. This can lead to
2182 // white text on a white bg with certain themes.
2183 if ($has_bgc_stl && !$has_txt_stl) {
2184 $styledef .= "color: $text; ";
2185 }
691a2d25 2186 if (strlen($styledef) > 0){
2187 $divattary{"style"} = "\"$styledef\"";
2188 }
2189 }
2190 return $divattary;
2191}
a3daaaf3 2192
691a2d25 2193/**
8bd0068d 2194 * This is the main function and the one you should actually be calling.
2195 * There are several variables you should be aware of an which need
2196 * special description.
2197 *
2198 * Since the description is quite lengthy, see it here:
2199 * http://linux.duke.edu/projects/mini/htmlfilter/
2200 *
2201 * @param $body the string with HTML you wish to filter
2202 * @param $tag_list see description above
2203 * @param $rm_tags_with_content see description above
2204 * @param $self_closing_tags see description above
2205 * @param $force_tag_closing see description above
2206 * @param $rm_attnames see description above
2207 * @param $bad_attvals see description above
2208 * @param $add_attr_to_tag see description above
2209 * @param $message message object
2210 * @param $id message id
2211 * @return sanitized html safe to show on your pages.
2212 */
da2415c1 2213function sq_sanitize($body,
8bd0068d 2214 $tag_list,
2215 $rm_tags_with_content,
2216 $self_closing_tags,
2217 $force_tag_closing,
2218 $rm_attnames,
2219 $bad_attvals,
2220 $add_attr_to_tag,
2221 $message,
2222 $id,
2223 $mailbox
2224 ){
b583c3e8 2225 $me = 'sq_sanitize';
7d06541f 2226 $rm_tags = array_shift($tag_list);
691a2d25 2227 /**
8bd0068d 2228 * Normalize rm_tags and rm_tags_with_content.
2229 */
7d06541f 2230 @array_walk($tag_list, 'sq_casenormalize');
691a2d25 2231 @array_walk($rm_tags_with_content, 'sq_casenormalize');
2232 @array_walk($self_closing_tags, 'sq_casenormalize');
2233 /**
8bd0068d 2234 * See if tag_list is of tags to remove or tags to allow.
2235 * false means remove these tags
2236 * true means allow these tags
2237 */
691a2d25 2238 $curpos = 0;
2239 $open_tags = Array();
2dd879b8 2240 $trusted = "\n<!-- begin sanitized html -->\n";
691a2d25 2241 $skip_content = false;
bb8d0799 2242 /**
8bd0068d 2243 * Take care of netscape's stupid javascript entities like
2244 * &{alert('boo')};
2245 */
bb8d0799 2246 $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
691a2d25 2247
7d06541f 2248 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
691a2d25 2249 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
2250 $free_content = substr($body, $curpos, $lt-$curpos);
2251 /**
8bd0068d 2252 * Take care of <style>
2253 */
7d06541f 2254 if ($tagname == "style" && $tagtype == 1){
da2415c1 2255 list($free_content, $curpos) =
e60a299a 2256 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
7d06541f 2257 if ($free_content != FALSE){
2258 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2259 $trusted .= $free_content;
2260 $trusted .= sq_tagprint($tagname, false, 2);
2261 }
2262 continue;
691a2d25 2263 }
2264 if ($skip_content == false){
2265 $trusted .= $free_content;
691a2d25 2266 }
2267 if ($tagname != FALSE){
2268 if ($tagtype == 2){
2269 if ($skip_content == $tagname){
2270 /**
8bd0068d 2271 * Got to the end of tag we needed to remove.
2272 */
691a2d25 2273 $tagname = false;
2274 $skip_content = false;
2275 } else {
2276 if ($skip_content == false){
c828931c 2277 if ($tagname == "body"){
2278 $tagname = "div";
2dd879b8 2279 }
da2415c1 2280 if (isset($open_tags{$tagname}) &&
8bd0068d 2281 $open_tags{$tagname} > 0){
2dd879b8 2282 $open_tags{$tagname}--;
691a2d25 2283 } else {
2dd879b8 2284 $tagname = false;
691a2d25 2285 }
691a2d25 2286 }
2287 }
2288 } else {
2289 /**
8bd0068d 2290 * $rm_tags_with_content
2291 */
691a2d25 2292 if ($skip_content == false){
2293 /**
8bd0068d 2294 * See if this is a self-closing type and change
2295 * tagtype appropriately.
2296 */
691a2d25 2297 if ($tagtype == 1
8bd0068d 2298 && in_array($tagname, $self_closing_tags)){
2dd879b8 2299 $tagtype = 3;
691a2d25 2300 }
2301 /**
8bd0068d 2302 * See if we should skip this tag and any content
2303 * inside it.
2304 */
691a2d25 2305 if ($tagtype == 1 &&
8bd0068d 2306 in_array($tagname, $rm_tags_with_content)){
691a2d25 2307 $skip_content = $tagname;
2308 } else {
da2415c1 2309 if (($rm_tags == false
8bd0068d 2310 && in_array($tagname, $tag_list)) ||
2311 ($rm_tags == true &&
2312 !in_array($tagname, $tag_list))){
691a2d25 2313 $tagname = false;
2314 } else {
2dd879b8 2315 /**
8bd0068d 2316 * Convert body into div.
2317 */
2dd879b8 2318 if ($tagname == "body"){
2319 $tagname = "div";
da2415c1 2320 $attary = sq_body2div($attary, $mailbox,
8bd0068d 2321 $message, $id);
2dd879b8 2322 }
691a2d25 2323 if ($tagtype == 1){
2324 if (isset($open_tags{$tagname})){
2325 $open_tags{$tagname}++;
2326 } else {
2327 $open_tags{$tagname}=1;
2328 }
2329 }
2330 /**
8bd0068d 2331 * This is where we run other checks.
2332 */
691a2d25 2333 if (is_array($attary) && sizeof($attary) > 0){
2334 $attary = sq_fixatts($tagname,
8bd0068d 2335 $attary,
2336 $rm_attnames,
2337 $bad_attvals,
2338 $add_attr_to_tag,
2339 $message,
2340 $id,
2341 $mailbox
2342 );
691a2d25 2343 }
2344 }
2345 }
691a2d25 2346 }
2347 }
2348 if ($tagname != false && $skip_content == false){
2349 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2350 }
691a2d25 2351 }
2352 $curpos = $gt+1;
a3daaaf3 2353 }
691a2d25 2354 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2355 if ($force_tag_closing == true){
2356 foreach ($open_tags as $tagname=>$opentimes){
2357 while ($opentimes > 0){
2358 $trusted .= '</' . $tagname . '>';
2359 $opentimes--;
2360 }
2361 }
2362 $trusted .= "\n";
2363 }
2364 $trusted .= "<!-- end sanitized html -->\n";
2365 return $trusted;
2366}
451f74a2 2367
691a2d25 2368/**
8bd0068d 2369 * This is a wrapper function to call html sanitizing routines.
2370 *
2371 * @param $body the body of the message
2372 * @param $id the id of the message
c189a963 2373
2374 * @param $message
2375 * @param $mailbox
2376 * @param boolean $take_mailto_links When TRUE, converts mailto: links
2377 * into internal SM compose links
2378 * (optional; default = TRUE)
8bd0068d 2379 * @return a string with html safe to display in the browser.
2380 */
c189a963 2381function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2382
202bcbcc 2383 // require_once(SM_PATH . 'functions/url_parser.php'); // for $MailTo_PReg_Match
c189a963 2384
691a2d25 2385 global $attachment_common_show_images, $view_unsafe_images,
8bd0068d 2386 $has_unsafe_images;
691a2d25 2387 /**
8bd0068d 2388 * Don't display attached images in HTML mode.
2b665f28 2389 *
d0187bd6 2390 * SB: why?
8bd0068d 2391 */
691a2d25 2392 $attachment_common_show_images = false;
2393 $tag_list = Array(
8bd0068d 2394 false,
2395 "object",
2396 "meta",
2397 "html",
2398 "head",
2399 "base",
2400 "link",
2401 "frame",
2402 "iframe",
2403 "plaintext",
2404 "marquee"
2405 );
691a2d25 2406
2407 $rm_tags_with_content = Array(
8bd0068d 2408 "script",
2409 "applet",
2410 "embed",
2411 "title",
2412 "frameset",
0493ed11 2413 "xmp",
8bd0068d 2414 "xml"
2415 );
691a2d25 2416
2417 $self_closing_tags = Array(
8bd0068d 2418 "img",
2419 "br",
2420 "hr",
2421 "input",
2422 "outbind"
2423 );
691a2d25 2424
2dd879b8 2425 $force_tag_closing = true;
691a2d25 2426
2427 $rm_attnames = Array(
8bd0068d 2428 "/.*/" =>
2429 Array(
2430 "/target/i",
2431 "/^on.*/i",
2432 "/^dynsrc/i",
2433 "/^data.*/i",
2434 "/^lowsrc.*/i"
2435 )
2436 );
691a2d25 2437
2438 $secremoveimg = "../images/" . _("sec_remove_eng.png");
2439 $bad_attvals = Array(
8bd0068d 2440 "/.*/" =>
691a2d25 2441 Array(
0a6ec9b5 2442 "/^src|background/i" =>
8bd0068d 2443 Array(
691a2d25 2444 Array(
8bd0068d 2445 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2446 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2447 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
691a2d25 2448 ),
8bd0068d 2449 Array(
2450 "\\1$secremoveimg\\2",
2451 "\\1$secremoveimg\\2",
2452 "\\1$secremoveimg\\2",
8bd0068d 2453 )
2454 ),
0a6ec9b5 2455 "/^href|action/i" =>
8bd0068d 2456 Array(
0a6ec9b5 2457 Array(
8bd0068d 2458 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2459 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2460 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
0a6ec9b5 2461 ),
691a2d25 2462 Array(
8bd0068d 2463 "\\1#\\1",
2464 "\\1#\\1",
2465 "\\1#\\1"
02474e43 2466 )
8bd0068d 2467 ),
2468 "/^style/i" =>
2469 Array(
2470 Array(
5db90261 2471 "/\/\*.*\*\//",
8bd0068d 2472 "/expression/i",
2473 "/binding/i",
2474 "/behaviou*r/i",
2475 "/include-source/i",
2476 "/position\s*:\s*absolute/i",
1d935bc2 2477 "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
8bd0068d 2478 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2479 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2480 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2481 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2482 ),
2483 Array(
5db90261 2484 "",
8bd0068d 2485 "idiocy",
2486 "idiocy",
2487 "idiocy",
2488 "idiocy",
567dc524 2489 "idiocy",
5b4884be 2490 "url",
8bd0068d 2491 "url(\\1#\\1)",
2492 "url(\\1#\\1)",
2493 "url(\\1#\\1)",
8bd0068d 2494 "\\1:url(\\2#\\3)"
2495 )
691a2d25 2496 )
8bd0068d 2497 )
691a2d25 2498 );
9ebace19 2499
2500 // If there's no "view_unsafe_images" setting in the user's preferences,
2501 // turn unsafe images off by default.
5262d9a6 2502 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
2dd879b8 2503 $view_unsafe_images = false;
45071bd6 2504 }
9ebace19 2505
691a2d25 2506 if (!$view_unsafe_images){
2507 /**
8bd0068d 2508 * Remove any references to http/https if view_unsafe_images set
2509 * to false.
2510 */
02474e43 2511 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
8bd0068d 2512 '/^([\'\"])\s*https*:.*([\'\"])/si');
02474e43 2513 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
8bd0068d 2514 "\\1$secremoveimg\\1");
02474e43 2515 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
7ef0b415 2516 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
02474e43 2517 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
8bd0068d 2518 "url(\\1$secremoveimg\\1)");
691a2d25 2519 }
451f74a2 2520
691a2d25 2521 $add_attr_to_tag = Array(
8bd0068d 2522 "/^a$/i" =>
c25f2fbb 2523 Array('target'=>'"_blank"',
02474e43 2524 'title'=>'"'._("This external link will open in a new window").'"'
8bd0068d 2525 )
2526 );
da2415c1 2527 $trusted = sq_sanitize($body,
8bd0068d 2528 $tag_list,
2529 $rm_tags_with_content,
2530 $self_closing_tags,
2531 $force_tag_closing,
2532 $rm_attnames,
2533 $bad_attvals,
2534 $add_attr_to_tag,
2535 $message,
2536 $id,
2537 $mailbox
2538 );
567dc524 2539 if (strpos($trusted,$secremoveimg)){
691a2d25 2540 $has_unsafe_images = true;
da2415c1 2541 }
c189a963 2542
2543 // we want to parse mailto's in HTML output, change to SM compose links
2544 // this is a modified version of code from url_parser.php... but Marc is
2545 // right: we need a better filtering implementation; adding this randomly
2546 // here is not a great solution
2547 //
2548 if ($take_mailto_links) {
2549 // parseUrl($trusted); // this even parses URLs inside of tags... too aggressive
2550 global $MailTo_PReg_Match;
202bcbcc 2551 $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1) ;
c189a963 2552 if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2553 foreach ($regs[0] as $i => $mailto_before) {
2554 $mailto_params = $regs[10][$i];
2555 // get rid of any tailing quote since we have to add send_to to the end
2556 //
2557 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2558 $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2559 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2560 $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2561
2562 if ($regs[1][$i]) { //if there is an email addr before '?', we need to merge it with the params
2563 $to = 'to=' . $regs[1][$i];
2564 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
2565 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2566 else {
2567 if ($mailto_params) //already some params, append to them
2568 $mailto_params .= '&amp;' . $to;
2569 else
2570 $mailto_params .= '?' . $to;
2571 }
2572 }
2573
2574 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2575
2576 // we'll already have target=_blank, no need to allow comp_in_new
2577 // here (which would be a lot more work anyway)
2578 //
2579 global $compose_new_win;
2580 $temp_comp_in_new = $compose_new_win;
2581 $compose_new_win = 0;
2582 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2583 $compose_new_win = $temp_comp_in_new;
2584
2585 // remove <a href=" and anything after the next quote (we only
2586 // need the uri, not the link HTML) in compose uri
2587 //
2588 $comp_uri = substr($comp_uri, 9);
2589 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2590 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2591 }
2592 }
2593 }
2594
691a2d25 2595 return $trusted;
451f74a2 2596}
a4a70693 2597
da2415c1 2598/**
8bd0068d 2599 * function SendDownloadHeaders - send file to the browser
2600 *
2601 * Original Source: SM core src/download.php
2602 * moved here to make it available to other code, and separate
2603 * front end from back end functionality.
2604 *
2605 * @param string $type0 first half of mime type
2606 * @param string $type1 second half of mime type
2607 * @param string $filename filename to tell the browser for downloaded file
2608 * @param boolean $force whether to force the download dialog to pop
2609 * @param optional integer $filesize send the Content-Header and length to the browser
2610 * @return void
2611 */
02474e43 2612function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2613 global $languages, $squirrelmail_language;
cfffd60b 2614 $isIE = $isIE6plus = false;
02474e43 2615
2616 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2617
2618 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
8bd0068d 2619 strstr($HTTP_USER_AGENT, 'Opera') === false) {
cfffd60b 2620 $isIE = true;
02474e43 2621 }
2622
cfffd60b 2623 if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2624 ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2625 $isIE6plus = true;
02474e43 2626 }
2627
2628 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 2629 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
02474e43 2630 $filename =
8bd0068d 2631 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
02474e43 2632 } else {
2633 $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
2634 }
2635
2636 // A Pox on Microsoft and it's Internet Explorer!
2637 //
2638 // IE has lots of bugs with file downloads.
2639 // It also has problems with SSL. Both of these cause problems
2640 // for us in this function.
2641 //
2642 // See this article on Cache Control headers and SSL
2643 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2644 //
2645 // The best thing you can do for IE is to upgrade to the latest
2646 // version
2647 //set all the Cache Control Headers for IE
2648 if ($isIE) {
2649 $filename=rawurlencode($filename);
2650 header ("Pragma: public");
8bd0068d 2651 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
02474e43 2652 header ("Cache-Control: post-check=0, pre-check=0", false);
8bd0068d 2653 header ("Cache-Control: private");
02474e43 2654
2655 //set the inline header for IE, we'll add the attachment header later if we need it
2656 header ("Content-Disposition: inline; filename=$filename");
2657 }
2658
2659 if (!$force) {
2660 // Try to show in browser window
2661 header ("Content-Disposition: inline; filename=\"$filename\"");
2662 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2663 } else {
2664 // Try to pop up the "save as" box
2665
2666 // IE makes this hard. It pops up 2 save boxes, or none.
2667 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2668 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2669 // But, according to Microsoft, it is "RFC compliant but doesn't
2670 // take into account some deviations that allowed within the
2671 // specification." Doesn't that mean RFC non-compliant?
2672 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2673
2674 // all browsers need the application/octet-stream header for this
2675 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2676
2677 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2678 // Do not have quotes around filename, but that applied to
2679 // "attachment"... does it apply to inline too?
2680 header ("Content-Disposition: attachment; filename=\"$filename\"");
2681
cfffd60b 2682 if ($isIE && !$isIE6plus) {
02474e43 2683 // This combination seems to work mostly. IE 5.5 SP 1 has
2684 // known issues (see the Microsoft Knowledge Base)
2685
2686 // This works for most types, but doesn't work with Word files
2687 header ("Content-Type: application/download; name=\"$filename\"");
7e2ff844 2688 header ("Content-Type: application/force-download; name=\"$filename\"");
02474e43 2689 // These are spares, just in case. :-)
2690 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2691 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2692 //header("Content-Type: application/octet-stream; name=\"$filename\"");
7e2ff844 2693 } else if ($isIE) {
2694 // This is to prevent IE for MIME sniffing and auto open a file in IE
2695 header ("Content-Type: application/force-download; name=\"$filename\"");
02474e43 2696 } else {
2697 // another application/octet-stream forces download for Netscape
2698 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2699 }
2700 }
2701
2702 //send the content-length header if the calling function provides it
2703 if ($filesize > 0) {
2704 header("Content-Length: $filesize");
2705 }
07c49f57 2706
8d863f64 2707} // end fn SendDownloadHeaders
Unnamed repository; edit this file 'description' to name the repository.