functions/mime.php

   1 <?php
   2
   3 /**
   4 * mime.php
   5 *
   6 * Copyright (c) 1999-2005 The SquirrelMail Project Team
   7 * Licensed under the GNU GPL. For full terms see the file COPYING.
   8 *
   9 * This contains the functions necessary to detect and decode MIME
  10 * messages.
  11 *
  12 * @version $Id$
  13 * @package squirrelmail
  14 */
  15
  16 /** The typical includes... */
  17 require_once(SM_PATH . 'functions/imap.php');
  18 require_once(SM_PATH . 'functions/attachment_common.php');
  19 /** add sqm_baseuri()*/
  20 include_once(SM_PATH . 'functions/display_messages.php');
  21
  22 /* -------------------------------------------------------------------------- */
  23 /* MIME DECODING                                                              */
  24 /* -------------------------------------------------------------------------- */
  25
  26 /**
  27 * Get the MIME structure
  28 *
  29 * This function gets the structure of a message and stores it in the "message" class.
  30 * It will return this object for use with all relevant header information and
  31 * fully parsed into the standard "message" object format.
  32 */
  33 function mime_structure ($bodystructure, $flags=array()) {
  34
  35     /* Isolate the body structure and remove beginning and end parenthesis. */
  36     $read = trim(substr ($bodystructure, strpos(strtolower($bodystructure), 'bodystructure') + 13));
  37     $read = trim(substr ($read, 0, -1));
  38     $i = 0;
  39     $msg = Message::parseStructure($read,$i);
  40     if (!is_object($msg)) {
  41         include_once(SM_PATH . 'functions/display_messages.php');
  42         global $color, $mailbox;
  43         /* removed urldecode because $_GET is auto urldecoded ??? */
  44         displayPageHeader( $color, $mailbox );
  45         $errormessage  = _("SquirrelMail could not decode the bodystructure of the message");
  46         $errormessage .= '<br />'._("the provided bodystructure by your imap-server").':<br /><br />';
  47         $errormessage .= '<pre>' . htmlspecialchars($read) . '</pre>';
  48         plain_error_message( $errormessage, $color );
  49         echo '</body></html>';
  50         exit;
  51     }
  52     if (count($flags)) {
  53         foreach ($flags as $flag) {
  54             $char = strtoupper($flag{1});
  55             switch ($char) {
  56                 case 'S':
  57                     if (strtolower($flag) == '\\seen') {
  58                         $msg->is_seen = true;
  59                     }
  60                     break;
  61                 case 'A':
  62                     if (strtolower($flag) == '\\answered') {
  63                         $msg->is_answered = true;
  64                     }
  65                     break;
  66                 case 'D':
  67                     if (strtolower($flag) == '\\deleted') {
  68                         $msg->is_deleted = true;
  69                     }
  70                     break;
  71                 case 'F':
  72                     if (strtolower($flag) == '\\flagged') {
  73                         $msg->is_flagged = true;
  74                     }
  75                     break;
  76                 case 'M':
  77                     if (strtolower($flag) == '$mdnsent') {
  78                         $msg->is_mdnsent = true;
  79                     }
  80                     break;
  81                 default:
  82                     break;
  83             }
  84         }
  85     }
  86     //    listEntities($msg);
  87     return $msg;
  88 }
  89
  90
  91
  92 /* This starts the parsing of a particular structure.  It is called recursively,
  93 * so it can be passed different structures.  It returns an object of type
  94 * $message.
  95 * First, it checks to see if it is a multipart message.  If it is, then it
  96 * handles that as it sees is necessary.  If it is just a regular entity,
  97 * then it parses it and adds the necessary header information (by calling out
  98 * to mime_get_elements()
  99 */
 100
 101 function mime_fetch_body($imap_stream, $id, $ent_id=1, $fetch_size=0) {
 102     /* Do a bit of error correction.  If we couldn't find the entity id, just guess
 103     * that it is the first one.  That is usually the case anyway.
 104     */
 105
 106     if (!$ent_id) {
 107         $cmd = "FETCH $id BODY[]";
 108     } else {
 109         $cmd = "FETCH $id BODY[$ent_id]";
 110     }
 111
 112     if ($fetch_size!=0) $cmd .= "<0.$fetch_size>";
 113
 114     $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, TRUE);
 115     do {
 116         $topline = trim(array_shift($data));
 117     } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
 118
 119     $wholemessage = implode('', $data);
 120     if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
 121         $ret = substr($wholemessage, 0, $regs[1]);
 122         /* There is some information in the content info header that could be important
 123         * in order to parse html messages. Let's get them here.
 124         */
 125 //        if ($ret{0} == '<') {
 126 //            $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, TRUE);
 127 //        }
 128     } else if (ereg('"([^"]*)"', $topline, $regs)) {
 129         $ret = $regs[1];
 130     } else {
 131         global $where, $what, $mailbox, $passed_id, $startMessage;
 132         $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
 133         if (isset($where) && isset($what)) {
 134             $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
 135         } else {
 136             $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 137         }
 138         $par .= '&amp;response=' . urlencode($response) .
 139                 '&amp;message='  . urlencode($message)  .
 140                 '&amp;topline='  . urlencode($topline);
 141
 142         echo   '<tt><br />' .
 143             '<table width="80%"><tr>' .
 144             '<tr><td colspan="2">' .
 145             _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
 146             '</td></tr>' .
 147             '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
 148             '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
 149             '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
 150             '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
 151             "</table><br /></tt></font><hr />";
 152
 153         $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, TRUE);
 154         array_shift($data);
 155         $wholemessage = implode('', $data);
 156
 157         $ret = $wholemessage;
 158     }
 159     return $ret;
 160 }
 161
 162 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding) {
 163
 164     /* Don't kill the connection if the browser is over a dialup
 165     * and it would take over 30 seconds to download it.
 166     * Don't call set_time_limit in safe mode.
 167     */
 168
 169     if (!ini_get('safe_mode')) {
 170         set_time_limit(0);
 171     }
 172     /* in case of base64 encoded attachments, do not buffer them.
 173     Instead, echo the decoded attachment directly to screen */
 174     if (strtolower($encoding) == 'base64') {
 175         if (!$ent_id) {
 176         $query = "FETCH $id BODY[]";
 177         } else {
 178         $query = "FETCH $id BODY[$ent_id]";
 179         }
 180         sqimap_run_command($imap_stream,$query,true,$response,$message,TRUE,'sqimap_base64_decode','php://stdout',true);
 181     } else {
 182     $body = mime_fetch_body ($imap_stream, $id, $ent_id);
 183     echo decodeBody($body, $encoding);
 184     }
 185
 186     /*
 187     TODO, use the same method for quoted printable.
 188     However, I assume that quoted printable attachments aren't that large
 189     so the performancegain / memory usage drop will be minimal.
 190     If we decide to add that then we need to adapt sqimap_fread because
 191     we need to split te result on \n and fread doesn't stop at \n. That
 192     means we also should provide $results from sqimap_fread (by ref) to
 193     te function and set $no_return to false. The $filter function for
 194     quoted printable should handle unsetting of $results.
 195     */
 196     /*
 197     TODO 2: find out how we write to the output stream php://stdout. fwrite
 198     doesn't work because 'php://stdout isn't a stream.
 199     */
 200
 201     return;
 202 /*
 203     fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
 204     $cnt = 0;
 205     $continue = true;
 206     $read = fgets ($imap_stream,8192);
 207
 208
 209     // This could be bad -- if the section has sqimap_session_id() . ' OK'
 210     // or similar, it will kill the download.
 211     while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
 212         if (trim($read) == ')==') {
 213             $read1 = $read;
 214             $read = fgets ($imap_stream,4096);
 215             if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
 216                 return;
 217             } else {
 218                 echo decodeBody($read1, $encoding) .
 219                     decodeBody($read, $encoding);
 220             }
 221         } else if ($cnt) {
 222             echo decodeBody($read, $encoding);
 223         }
 224         $read = fgets ($imap_stream,4096);
 225         $cnt++;
 226 //      break;
 227     }
 228 */
 229 }
 230
 231 /* -[ END MIME DECODING ]----------------------------------------------------------- */
 232
 233 /* This is here for debugging purposes.  It will print out a list
 234 * of all the entity IDs that are in the $message object.
 235 */
 236 function listEntities ($message) {
 237     if ($message) {
 238         echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br />';
 239         for ($i = 0; isset($message->entities[$i]); $i++) {
 240             echo "$i : ";
 241             $msg = listEntities($message->entities[$i]);
 242
 243             if ($msg) {
 244                 echo "return: ";
 245                 return $msg;
 246             }
 247         }
 248     }
 249 }
 250
 251 function getPriorityStr($priority) {
 252     $priority_level = substr($priority,0,1);
 253
 254     switch($priority_level) {
 255         /* Check for a higher then normal priority. */
 256         case '1':
 257         case '2':
 258             $priority_string = _("High");
 259             break;
 260
 261         /* Check for a lower then normal priority. */
 262         case '4':
 263         case '5':
 264             $priority_string = _("Low");
 265             break;
 266
 267         /* Check for a normal priority. */
 268         case '3':
 269         default:
 270             $priority_level = '3';
 271             $priority_string = _("Normal");
 272             break;
 273
 274     }
 275     return $priority_string;
 276 }
 277
 278 /* returns a $message object for a particular entity id */
 279 function getEntity ($message, $ent_id) {
 280     return $message->getEntity($ent_id);
 281 }
 282
 283 /* translateText
 284 * Extracted from strings.php 23/03/2002
 285 */
 286
 287 function translateText(&$body, $wrap_at, $charset) {
 288     global $where, $what;   /* from searching */
 289     global $color;          /* color theme */
 290
 291     require_once(SM_PATH . 'functions/url_parser.php');
 292
 293     $body_ary = explode("\n", $body);
 294     for ($i=0; $i < count($body_ary); $i++) {
 295         $line = $body_ary[$i];
 296         if (strlen($line) - 2 >= $wrap_at) {
 297             sqWordWrap($line, $wrap_at, $charset);
 298         }
 299         $line = charset_decode($charset, $line);
 300         $line = str_replace("\t", '        ', $line);
 301
 302         parseUrl ($line);
 303
 304         $quotes = 0;
 305         $pos = 0;
 306         $j = strlen($line);
 307
 308         while ($pos < $j) {
 309             if ($line[$pos] == ' ') {
 310                 $pos++;
 311             } else if (strpos($line, '&gt;', $pos) === $pos) {
 312                 $pos += 4;
 313                 $quotes++;
 314             } else {
 315                 break;
 316             }
 317         }
 318
 319         if ($quotes % 2) {
 320             if (!isset($color[13])) {
 321                 $color[13] = '#800000';
 322             }
 323             $line = '<font color="' . $color[13] . '">' . $line . '</font>';
 324         } elseif ($quotes) {
 325             if (!isset($color[14])) {
 326                 $color[14] = '#FF0000';
 327             }
 328             $line = '<font color="' . $color[14] . '">' . $line . '</font>';
 329         }
 330
 331         $body_ary[$i] = $line;
 332     }
 333     $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
 334 }
 335
 336 /**
 337  * This returns a parsed string called $body. That string can then
 338  * be displayed as the actual message in the HTML. It contains
 339  * everything needed, including HTML Tags, Attachments at the
 340  * bottom, etc.
 341  *
 342  * Since 1.2.0 function uses message_body hook.
 343  * Till 1.3.0 function included output of formatAttachments().
 344  *
 345  * @param resource $imap_stream imap connection resource
 346  * @param object $message squirrelmail message object
 347  * @param array $color squirrelmail color theme array
 348  * @param integer $wrap_at number of characters per line
 349  * @param string $ent_num (since 1.3.0) message part id
 350  * @param integer $id (since 1.3.0) message id
 351  * @param string $mailbox (since 1.3.0) imap folder name
 352  * @param boolean $clean (since 1.5.1) Do not output stuff that's irrelevant for the printable version.
 353  * @return string html formated message text
 354  */
 355 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX', $clean=FALSE) {
 356     /* This if statement checks for the entity to show as the
 357     * primary message. To add more of them, just put them in the
 358     * order that is their priority.
 359     */
 360     global $startMessage, $languages, $squirrelmail_language,
 361         $show_html_default, $sort, $has_unsafe_images, $passed_ent_id, $use_iframe,$iframe_height;
 362
 363     // workaround for not updated config.php
 364     if (! isset($use_iframe)) $use_iframe = false;
 365
 366     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
 367         $view_unsafe_images = false;
 368     }
 369
 370     $body = '';
 371     $urlmailbox = urlencode($mailbox);
 372     $body_message = getEntity($message, $ent_num);
 373     if (($body_message->header->type0 == 'text') ||
 374         ($body_message->header->type0 == 'rfc822')) {
 375         $body = mime_fetch_body ($imap_stream, $id, $ent_num);
 376         $body = decodeBody($body, $body_message->header->encoding);
 377
 378         if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 379             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
 380             if (mb_detect_encoding($body) != 'ASCII') {
 381                 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
 382             }
 383         }
 384         $hookResults = do_hook("message_body", $body);
 385         $body = $hookResults[1];
 386
 387         /* If there are other types that shouldn't be formatted, add
 388         * them here.
 389         */
 390
 391         if ($body_message->header->type1 == 'html') {
 392             if ($show_html_default <> 1) {
 393                 $entity_conv = array('&nbsp;' => ' ',
 394                                     '<p>'    => "\n",
 395                                     '<P>'    => "\n",
 396                                     '<br>'   => "\n",
 397                                     '<BR>'   => "\n",
 398                                     '<br />' => "\n",
 399                                     '<BR />' => "\n",
 400                                     '&gt;'   => '>',
 401                                     '&lt;'   => '<');
 402                 $body = strtr($body, $entity_conv);
 403                 $body = strip_tags($body);
 404                 $body = trim($body);
 405                 translateText($body, $wrap_at,
 406                             $body_message->header->getParameter('charset'));
 407             } elseif ($use_iframe && ! $clean) {
 408                 // $clean is used to remove iframe in printable view.
 409
 410                 /**
 411                  * If we don't add html message between iframe tags,
 412                  * we must detect unsafe images and modify $has_unsafe_images.
 413                  */
 414                 $html_body =  magicHTML($body, $id, $message, $mailbox);
 415
 416                 // creating iframe url
 417                 $iframeurl=sqm_baseuri().'src/view_html.php?'
 418                     . 'mailbox=' . $urlmailbox
 419                     . '&amp;passed_id=' . $id
 420                     . '&amp;ent_id=' . $ent_num
 421                     . '&amp;view_unsafe_images=' . (int) $view_unsafe_images;
 422
 423                 // adding warning message
 424                 $body = html_tag('div',_("Viewing HTML formatted email"),'center');
 425
 426                 /**
 427                  * height can't be set to 100%, because it does not work as expected when
 428                  * iframe is inside the table. Browsers do not create full height objects
 429                  * even when iframe is not nested. Maybe there is some way to get full size
 430                  * with CSS. Tested in firefox 1.02 and opera 7.53
 431                  *
 432                  * width="100%" does not work as expected, when table width is not set (automatic)
 433                  *
 434                  * tokul: I think <iframe> are safer sandbox than <object>. Objects might
 435                  * need special handling for IE and IE6SP2.
 436                  */
 437                 $body.= "<div><iframe name=\"message_frame\" width=\"100%\" height=\"$iframe_height\" src=\"$iframeurl\""
 438                     .' frameborder="1" marginwidth="0" marginheight="0" scrolling="auto">' . "\n";
 439
 440                 // Message for browsers without iframe support
 441                 //$body.= _("Your browser does not support inline frames.
 442                 // You can view HTML formated message by following below link.");
 443                 //$body.= "<br /><a href=\"$iframeurl\">"._("View HTML Message")."</a>";
 444
 445                 // if browser can't render iframe, it renders html message.
 446                 $body.= $html_body;
 447
 448                 // close iframe
 449                 $body.="</iframe></div>\n";
 450             } else {
 451                 // old way of html rendering
 452                 $body = magicHTML($body, $id, $message, $mailbox);
 453             }
 454         } else {
 455             translateText($body, $wrap_at,
 456                         $body_message->header->getParameter('charset'));
 457         }
 458
 459         // if this is the clean display (i.e. printer friendly), stop here.
 460         if ( $clean ) {
 461             return $body;
 462         }
 463
 464         $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
 465                 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
 466                 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 467         if (isset($passed_ent_id)) {
 468             $link .= '&amp;passed_ent_id='.$passed_ent_id;
 469         }
 470         $body .= '<center><small><a href="download.php?absolute_dl=true&amp;' .
 471                 $link . '">' . _("Download this as a file") .  '</a>';
 472         if ($view_unsafe_images) {
 473             $text = _("Hide Unsafe Images");
 474         } else {
 475             if (isset($has_unsafe_images) && $has_unsafe_images) {
 476                 $link .= '&amp;view_unsafe_images=1';
 477                 $text = _("View Unsafe Images");
 478             } else {
 479                 $text = '';
 480             }
 481         }
 482         if($text != '') {
 483             $body .= '&nbsp;|&nbsp;<a href="read_body.php?' . $link . '">' . $text . '</a>';
 484         }
 485         $body .= '</small></center><br />' . "\n";
 486     }
 487     return $body;
 488 }
 489
 490 /**
 491  * Displays attachment links and information
 492  *
 493  * Since 1.3.0 function is not included in formatBody() call.
 494  *
 495  * Since 1.0.2 uses attachment $type0/$type1 hook.
 496  * Since 1.2.5 uses attachment $type0/* hook.
 497  * Since 1.5.0 uses attachments_bottom hook.
 498  *
 499  * @param object $message SquirrelMail message object
 500  * @param array $exclude_id message parts that are not attachments.
 501  * @param string $mailbox mailbox name
 502  * @param integer $id message id
 503  * @return string html formated attachment information.
 504  */
 505 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 506     global $where, $what, $startMessage, $color, $passed_ent_id;
 507
 508     $att_ar = $message->getAttachments($exclude_id);
 509
 510     if (!count($att_ar)) return '';
 511
 512     $attachments = '';
 513
 514     $urlMailbox = urlencode($mailbox);
 515
 516     foreach ($att_ar as $att) {
 517         $ent = $att->entity_id;
 518         $header = $att->header;
 519         $type0 = strtolower($header->type0);
 520         $type1 = strtolower($header->type1);
 521         $name = '';
 522         $links['download link']['text'] = _("Download");
 523         $links['download link']['href'] = sqm_baseuri() .
 524                 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 525         if ($type0 =='message' && $type1 == 'rfc822') {
 526             $default_page = sqm_baseuri() . 'src/read_body.php';
 527             $rfc822_header = $att->rfc822_header;
 528             $filename = $rfc822_header->subject;
 529             if (trim( $filename ) == '') {
 530                 $filename = 'untitled-[' . $ent . ']' ;
 531             }
 532             $from_o = $rfc822_header->from;
 533             if (is_object($from_o)) {
 534                 $from_name = decodeHeader($from_o->getAddress(false));
 535             } else {
 536                 $from_name = _("Unknown sender");
 537             }
 538             $description = $from_name;
 539         } else {
 540             $default_page = sqm_baseuri() . 'src/download.php';
 541             $filename = $att->getFilename();
 542             if ($header->description) {
 543                 $description = decodeHeader($header->description);
 544             } else {
 545                 $description = '';
 546             }
 547         }
 548
 549         $display_filename = $filename;
 550         if (isset($passed_ent_id)) {
 551             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 552         } else {
 553             $passed_ent_id_link = '';
 554         }
 555         $defaultlink = $default_page . "?startMessage=$startMessage"
 556                     . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 557                     . '&amp;ent_id='.$ent.$passed_ent_id_link;
 558         if ($where && $what) {
 559         $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 560         }
 561
 562         /* This executes the attachment hook with a specific MIME-type.
 563         * If that doesn't have results, it tries if there's a rule
 564         * for a more generic type.
 565         */
 566         $hookresults = do_hook("attachment $type0/$type1", $links,
 567                             $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 568                             $display_filename, $where, $what);
 569         if(count($hookresults[1]) <= 1) {
 570             $hookresults = do_hook("attachment $type0/*", $links,
 571                                 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 572                                 $display_filename, $where, $what);
 573         }
 574
 575         $links = $hookresults[1];
 576         $defaultlink = $hookresults[6];
 577
 578         $attachments .= '<tr><td>' .
 579                         '<a href="'.$defaultlink.'">'.decodeHeader($display_filename).'</a>&nbsp;</td>' .
 580                         '<td><small><b>' . show_readable_size($header->size) .
 581                         '</b>&nbsp;&nbsp;</small></td>' .
 582                         '<td><small>[ '.htmlspecialchars($type0).'/'.htmlspecialchars($type1).' ]&nbsp;</small></td>' .
 583                         '<td><small>';
 584         $attachments .= '<b>' . $description . '</b>';
 585         $attachments .= '</small></td><td><small>&nbsp;';
 586
 587         $skipspaces = 1;
 588         foreach ($links as $val) {
 589             if ($skipspaces) {
 590                 $skipspaces = 0;
 591             } else {
 592                 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
 593             }
 594             $attachments .= '<a href="' . $val['href'] . '">'
 595                 . (isset($val['text']) && !empty($val['text']) ? $val['text'] : '')
 596                 . (isset($val['extra']) && !empty($val['extra']) ? $val['extra'] : '')
 597                 . '</a>';
 598         }
 599         unset($links);
 600         $attachments .= "</td></tr>\n";
 601     }
 602     $attachmentadd = do_hook_function('attachments_bottom',$attachments);
 603     if ($attachmentadd != '')
 604         $attachments = $attachmentadd;
 605     return $attachments;
 606 }
 607
 608 function sqimap_base64_decode(&$string) {
 609
 610     // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
 611     // fly decoding (to reduce memory usage) you have to check if the
 612     // data has incomplete pairs
 613
 614     // Remove the noise in order to check if the 4 bytes pairs are complete
 615     $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
 616
 617     $sStringRem = '';
 618     $iMod = strlen($string) % 4;
 619     if ($iMod) {
 620         $sStringRem = substr($string,-$iMod);
 621         // Check if $sStringRem contains padding characters
 622         if (substr($sStringRem,-1) != '=') {
 623             $string = substr($string,0,-$iMod);
 624         } else {
 625             $sStringRem = '';
 626         }
 627     }
 628     $string = base64_decode($string);
 629     return $sStringRem;
 630 }
 631
 632
 633 /* This function decodes the body depending on the encoding type. */
 634 function decodeBody($body, $encoding) {
 635     global $show_html_default;
 636
 637     $body = str_replace("\r\n", "\n", $body);
 638     $encoding = strtolower($encoding);
 639
 640     $encoding_handler = do_hook_function('decode_body', $encoding);
 641
 642
 643     // plugins get first shot at decoding the body
 644     //
 645     if (!empty($encoding_handler) && function_exists($encoding_handler)) {
 646         $body = $encoding_handler('decode', $body);
 647
 648     } else if ($encoding == 'quoted-printable' ||
 649         $encoding == 'quoted_printable') {
 650         $body = quoted_printable_decode($body);
 651
 652         while (ereg("=\n", $body)) {
 653             $body = ereg_replace ("=\n", '', $body);
 654         }
 655
 656     } else if ($encoding == 'base64') {
 657         $body = base64_decode($body);
 658     }
 659
 660     // All other encodings are returned raw.
 661     return $body;
 662 }
 663
 664 /**
 665 * Decodes headers
 666 *
 667 * This functions decode strings that is encoded according to
 668 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 669 * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 670 *
 671 * @param string $string header string that has to be made readable
 672 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
 673 * @param boolean $htmlsave preserve spaces and sanitize html special characters. defaults to true
 674 * @param boolean $decide decide if string can be utfencoded. defaults to false
 675 * @return string decoded header string
 676 */
 677 function decodeHeader ($string, $utfencode=true,$htmlsave=true,$decide=false) {
 678     global $languages, $squirrelmail_language,$default_charset;
 679     if (is_array($string)) {
 680         $string = implode("\n", $string);
 681     }
 682
 683     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 684         function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
 685         $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
 686         // Do we need to return at this point?
 687         // return $string;
 688     }
 689     $i = 0;
 690     $iLastMatch = -2;
 691     $encoded = true;
 692
 693     $aString = explode(' ',$string);
 694     $ret = '';
 695     foreach ($aString as $chunk) {
 696         if ($encoded && $chunk === '') {
 697             continue;
 698         } elseif ($chunk === '') {
 699             $ret .= ' ';
 700             continue;
 701         }
 702         $encoded = false;
 703         /* if encoded words are not separated by a linear-space-white we still catch them */
 704         $j = $i-1;
 705
 706         while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 707             /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
 708             if ($iLastMatch !== $j) {
 709                 if ($htmlsave) {
 710                     $ret .= '&#32;';
 711                 } else {
 712                     $ret .= ' ';
 713                 }
 714             }
 715             $iLastMatch = $i;
 716             $j = $i;
 717             if ($htmlsave) {
 718                 $ret .= htmlspecialchars($res[1]);
 719             } else {
 720                 $ret .= $res[1];
 721             }
 722             $encoding = ucfirst($res[3]);
 723
 724             /* decide about valid decoding */
 725             if ($decide && is_conversion_safe($res[2])) {
 726             $utfencode=true;
 727             $can_be_encoded=true;
 728             } else {
 729             $can_be_encoded=false;
 730             }
 731             switch ($encoding)
 732             {
 733             case 'B':
 734                 $replace = base64_decode($res[4]);
 735                 if ($utfencode) {
 736                     if ($can_be_encoded) {
 737                         /* convert string to different charset,
 738                         * if functions asks for it (usually in compose)
 739                         */
 740                         $ret .= charset_convert($res[2],$replace,$default_charset);
 741                     } else {
 742                         // convert string to html codes in order to display it
 743                         $ret .= charset_decode($res[2],$replace);
 744                     }
 745                 } else {
 746                     if ($htmlsave) {
 747                         $replace = htmlspecialchars($replace);
 748                     }
 749                     $ret.= $replace;
 750                 }
 751                 break;
 752             case 'Q':
 753                 $replace = str_replace('_', ' ', $res[4]);
 754                 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 755                                     $replace);
 756                 if ($utfencode) {
 757                 if ($can_be_encoded) {
 758                     /* convert string to different charset,
 759                     * if functions asks for it (usually in compose)
 760                     */
 761                     $replace = charset_convert($res[2], $replace,$default_charset);
 762                 } else {
 763                     // convert string to html codes in order to display it
 764                     $replace = charset_decode($res[2], $replace);
 765                 }
 766                 } else {
 767                     if ($htmlsave) {
 768                         $replace = htmlspecialchars($replace);
 769                     }
 770                 }
 771                 $ret .= $replace;
 772                 break;
 773             default:
 774                 break;
 775             }
 776             $chunk = $res[5];
 777             $encoded = true;
 778         }
 779         if (!$encoded) {
 780             if ($htmlsave) {
 781                 $ret .= '&#32;';
 782             } else {
 783                 $ret .= ' ';
 784             }
 785         }
 786
 787         if (!$encoded && $htmlsave) {
 788             $ret .= htmlspecialchars($chunk);
 789         } else {
 790             $ret .= $chunk;
 791         }
 792         ++$i;
 793     }
 794     /* remove the first added space */
 795     if ($ret) {
 796         if ($htmlsave) {
 797             $ret = substr($ret,5);
 798         } else {
 799             $ret = substr($ret,1);
 800         }
 801     }
 802
 803     return $ret;
 804 }
 805
 806 /**
 807 * Encodes header as quoted-printable
 808 *
 809 * Encode a string according to RFC 1522 for use in headers if it
 810 * contains 8-bit characters or anything that looks like it should
 811 * be encoded.
 812 *
 813 * @param string $string header string, that has to be encoded
 814 * @return string quoted-printable encoded string
 815 */
 816 function encodeHeader ($string) {
 817     global $default_charset, $languages, $squirrelmail_language;
 818
 819     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 820         function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
 821         return  call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
 822     }
 823
 824     // Encode only if the string contains 8-bit characters or =?
 825     $j = strlen($string);
 826     $max_l = 75 - strlen($default_charset) - 7;
 827     $aRet = array();
 828     $ret = '';
 829     $iEncStart = $enc_init = false;
 830     $cur_l = $iOffset = 0;
 831     for($i = 0; $i < $j; ++$i) {
 832         switch($string{$i})
 833         {
 834         case '=':
 835         case '<':
 836         case '>':
 837         case ',':
 838         case '?':
 839         case '_':
 840             if ($iEncStart === false) {
 841                 $iEncStart = $i;
 842             }
 843             $cur_l+=3;
 844             if ($cur_l > ($max_l-2)) {
 845                 /* if there is an stringpart that doesn't need encoding, add it */
 846                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 847                 $aRet[] = "=?$default_charset?Q?$ret?=";
 848                 $iOffset = $i;
 849                 $cur_l = 0;
 850                 $ret = '';
 851                 $iEncStart = false;
 852             } else {
 853                 $ret .= sprintf("=%02X",ord($string{$i}));
 854             }
 855             break;
 856         case '(':
 857         case ')':
 858             if ($iEncStart !== false) {
 859                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 860                 $aRet[] = "=?$default_charset?Q?$ret?=";
 861                 $iOffset = $i;
 862                 $cur_l = 0;
 863                 $ret = '';
 864                 $iEncStart = false;
 865             }
 866             break;
 867         case ' ':
 868             if ($iEncStart !== false) {
 869                 $cur_l++;
 870                 if ($cur_l > $max_l) {
 871                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 872                     $aRet[] = "=?$default_charset?Q?$ret?=";
 873                     $iOffset = $i;
 874                     $cur_l = 0;
 875                     $ret = '';
 876                     $iEncStart = false;
 877                 } else {
 878                     $ret .= '_';
 879                 }
 880             }
 881             break;
 882         default:
 883             $k = ord($string{$i});
 884             if ($k > 126) {
 885                 if ($iEncStart === false) {
 886                     // do not start encoding in the middle of a string, also take the rest of the word.
 887                     $sLeadString = substr($string,0,$i);
 888                     $aLeadString = explode(' ',$sLeadString);
 889                     $sToBeEncoded = array_pop($aLeadString);
 890                     $iEncStart = $i - strlen($sToBeEncoded);
 891                     $ret .= $sToBeEncoded;
 892                     $cur_l += strlen($sToBeEncoded);
 893                 }
 894                 $cur_l += 3;
 895                 /* first we add the encoded string that reached it's max size */
 896                 if ($cur_l > ($max_l-2)) {
 897                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 898                     $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
 899                     $cur_l = 3;
 900                     $ret = '';
 901                     $iOffset = $i;
 902                     $iEncStart = $i;
 903                 }
 904                 $enc_init = true;
 905                 $ret .= sprintf("=%02X", $k);
 906             } else {
 907                 if ($iEncStart !== false) {
 908                     $cur_l++;
 909                     if ($cur_l > $max_l) {
 910                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 911                         $aRet[] = "=?$default_charset?Q?$ret?=";
 912                         $iEncStart = false;
 913                         $iOffset = $i;
 914                         $cur_l = 0;
 915                         $ret = '';
 916                     } else {
 917                         $ret .= $string{$i};
 918                     }
 919                 }
 920             }
 921             break;
 922         }
 923     }
 924
 925     if ($enc_init) {
 926         if ($iEncStart !== false) {
 927             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 928             $aRet[] = "=?$default_charset?Q?$ret?=";
 929         } else {
 930             $aRet[] = substr($string,$iOffset);
 931         }
 932         $string = implode('',$aRet);
 933     }
 934     return $string;
 935 }
 936
 937 /* This function trys to locate the entity_id of a specific mime element */
 938 function find_ent_id($id, $message) {
 939     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
 940         if ($message->entities[$i]->header->type0 == 'multipart')  {
 941             $ret = find_ent_id($id, $message->entities[$i]);
 942         } else {
 943             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
 944 //                if (sq_check_save_extension($message->entities[$i])) {
 945                     return $message->entities[$i]->entity_id;
 946 //                }
 947             } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
 948                 /**
 949                  * This is part of a fix for Outlook Express 6.x generating
 950                  * cid URLs without creating content-id headers
 951                  * @@JA - 20050207
 952                  */
 953                 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
 954                     return $message->entities[$i]->entity_id;
 955                 }
 956             }
 957         }
 958     }
 959     return $ret;
 960 }
 961
 962 function sq_check_save_extension($message) {
 963     $filename = $message->getFilename();
 964     $ext = substr($filename, strrpos($filename,'.')+1);
 965     $save_extensions = array('jpg','jpeg','gif','png','bmp');
 966     return in_array($ext, $save_extensions);
 967 }
 968
 969
 970 /**
 971 ** HTMLFILTER ROUTINES
 972 */
 973
 974 /**
 975 * This function is more or less a wrapper around stripslashes. Apparently
 976 * Explorer is stupid enough to just remove the backslashes and then
 977 * execute the content of the attribute as if nothing happened.
 978 * Who does that?
 979 *
 980 * @param  attvalue   The value of the attribute
 981 * @return attvalue   The value of the attribute stripslashed.
 982 */
 983 function sq_unbackslash($attvalue){
 984     /**
 985     * Remove any backslashes. See if there are any first.
 986     */
 987
 988     if (strstr($attvalue, '\\') !== false){
 989         $attvalue = stripslashes($attvalue);
 990     }
 991     return $attvalue;
 992 }
 993
 994 /**
 995 * Kill any tabs, newlines, or carriage returns. Our friends the
 996 * makers of the browser with 95% market value decided that it'd
 997 * be funny to make "java[tab]script" be just as good as "javascript".
 998 *
 999 * @param  attvalue  The attribute value before extraneous spaces removed.
1000 * @return attvalue  The attribute value after extraneous spaces removed.
1001 */
1002 function sq_unspace($attvalue){
1003     if (strcspn($attvalue, "\t\r\n") != strlen($attvalue)){
1004         $attvalue = str_replace(Array("\t", "\r", "\n"), Array('', '', ''),
1005                                 $attvalue);
1006     }
1007     return $attvalue;
1008 }
1009
1010 /**
1011 * This function returns the final tag out of the tag name, an array
1012 * of attributes, and the type of the tag. This function is called by
1013 * sq_sanitize internally.
1014 *
1015 * @param  $tagname  the name of the tag.
1016 * @param  $attary   the array of attributes and their values
1017 * @param  $tagtype  The type of the tag (see in comments).
1018 * @return           a string with the final tag representation.
1019 */
1020 function sq_tagprint($tagname, $attary, $tagtype){
1021     $me = 'sq_tagprint';
1022
1023     if ($tagtype == 2){
1024         $fulltag = '</' . $tagname . '>';
1025     } else {
1026         $fulltag = '<' . $tagname;
1027         if (is_array($attary) && sizeof($attary)){
1028             $atts = Array();
1029             while (list($attname, $attvalue) = each($attary)){
1030                 array_push($atts, "$attname=$attvalue");
1031             }
1032             $fulltag .= ' ' . join(" ", $atts);
1033         }
1034         if ($tagtype == 3){
1035             $fulltag .= ' /';
1036         }
1037         $fulltag .= '>';
1038     }
1039     return $fulltag;
1040 }
1041
1042 /**
1043 * A small helper function to use with array_walk. Modifies a by-ref
1044 * value and makes it lowercase.
1045 *
1046 * @param  $val a value passed by-ref.
1047 * @return      void since it modifies a by-ref value.
1048 */
1049 function sq_casenormalize(&$val){
1050     $val = strtolower($val);
1051 }
1052
1053 /**
1054 * This function skips any whitespace from the current position within
1055 * a string and to the next non-whitespace value.
1056 *
1057 * @param  $body   the string
1058 * @param  $offset the offset within the string where we should start
1059 *                 looking for the next non-whitespace character.
1060 * @return         the location within the $body where the next
1061 *                 non-whitespace char is located.
1062 */
1063 function sq_skipspace($body, $offset){
1064     $me = 'sq_skipspace';
1065     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1066     if (sizeof($matches{1})){
1067         $count = strlen($matches{1});
1068         $offset += $count;
1069     }
1070     return $offset;
1071 }
1072
1073 /**
1074 * This function looks for the next character within a string.  It's
1075 * really just a glorified "strpos", except it catches if failures
1076 * nicely.
1077 *
1078 * @param  $body   The string to look for needle in.
1079 * @param  $offset Start looking from this position.
1080 * @param  $needle The character/string to look for.
1081 * @return         location of the next occurance of the needle, or
1082 *                 strlen($body) if needle wasn't found.
1083 */
1084 function sq_findnxstr($body, $offset, $needle){
1085     $me  = 'sq_findnxstr';
1086     $pos = strpos($body, $needle, $offset);
1087     if ($pos === FALSE){
1088         $pos = strlen($body);
1089     }
1090     return $pos;
1091 }
1092
1093 /**
1094 * This function takes a PCRE-style regexp and tries to match it
1095 * within the string.
1096 *
1097 * @param  $body   The string to look for needle in.
1098 * @param  $offset Start looking from here.
1099 * @param  $reg    A PCRE-style regex to match.
1100 * @return         Returns a false if no matches found, or an array
1101 *                 with the following members:
1102 *                 - integer with the location of the match within $body
1103 *                 - string with whatever content between offset and the match
1104 *                 - string with whatever it is we matched
1105 */
1106 function sq_findnxreg($body, $offset, $reg){
1107     $me = 'sq_findnxreg';
1108     $matches = Array();
1109     $retarr = Array();
1110     preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1111     if (!isset($matches{0}) || !$matches{0}){
1112         $retarr = false;
1113     } else {
1114         $retarr{0} = $offset + strlen($matches{1});
1115         $retarr{1} = $matches{1};
1116         $retarr{2} = $matches{2};
1117     }
1118     return $retarr;
1119 }
1120
1121 /**
1122 * This function looks for the next tag.
1123 *
1124 * @param  $body   String where to look for the next tag.
1125 * @param  $offset Start looking from here.
1126 * @return         false if no more tags exist in the body, or
1127 *                 an array with the following members:
1128 *                 - string with the name of the tag
1129 *                 - array with attributes and their values
1130 *                 - integer with tag type (1, 2, or 3)
1131 *                 - integer where the tag starts (starting "<")
1132 *                 - integer where the tag ends (ending ">")
1133 *                 first three members will be false, if the tag is invalid.
1134 */
1135 function sq_getnxtag($body, $offset){
1136     $me = 'sq_getnxtag';
1137     if ($offset > strlen($body)){
1138         return false;
1139     }
1140     $lt = sq_findnxstr($body, $offset, "<");
1141     if ($lt == strlen($body)){
1142         return false;
1143     }
1144     /**
1145     * We are here:
1146     * blah blah <tag attribute="value">
1147     * \---------^
1148     */
1149     $pos = sq_skipspace($body, $lt+1);
1150     if ($pos >= strlen($body)){
1151         return Array(false, false, false, $lt, strlen($body));
1152     }
1153     /**
1154     * There are 3 kinds of tags:
1155     * 1. Opening tag, e.g.:
1156     *    <a href="blah">
1157     * 2. Closing tag, e.g.:
1158     *    </a>
1159     * 3. XHTML-style content-less tag, e.g.:
1160     *    <img src="blah" />
1161     */
1162     $tagtype = false;
1163     switch (substr($body, $pos, 1)){
1164         case '/':
1165             $tagtype = 2;
1166             $pos++;
1167             break;
1168         case '!':
1169             /**
1170             * A comment or an SGML declaration.
1171             */
1172             if (substr($body, $pos+1, 2) == "--"){
1173                 $gt = strpos($body, "-->", $pos);
1174                 if ($gt === false){
1175                     $gt = strlen($body);
1176                 } else {
1177                     $gt += 2;
1178                 }
1179                 return Array(false, false, false, $lt, $gt);
1180             } else {
1181                 $gt = sq_findnxstr($body, $pos, ">");
1182                 return Array(false, false, false, $lt, $gt);
1183             }
1184             break;
1185         default:
1186             /**
1187             * Assume tagtype 1 for now. If it's type 3, we'll switch values
1188             * later.
1189             */
1190             $tagtype = 1;
1191             break;
1192     }
1193
1194     $tagname = '';
1195     /**
1196     * Look for next [\W-_], which will indicate the end of the tag name.
1197     */
1198     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1199     if ($regary == false){
1200         return Array(false, false, false, $lt, strlen($body));
1201     }
1202     list($pos, $tagname, $match) = $regary;
1203     $tagname = strtolower($tagname);
1204
1205     /**
1206     * $match can be either of these:
1207     * '>'  indicating the end of the tag entirely.
1208     * '\s' indicating the end of the tag name.
1209     * '/'  indicating that this is type-3 xhtml tag.
1210     *
1211     * Whatever else we find there indicates an invalid tag.
1212     */
1213     switch ($match){
1214         case '/':
1215             /**
1216             * This is an xhtml-style tag with a closing / at the
1217             * end, like so: <img src="blah" />. Check if it's followed
1218             * by the closing bracket. If not, then this tag is invalid
1219             */
1220             if (substr($body, $pos, 2) == "/>"){
1221                 $pos++;
1222                 $tagtype = 3;
1223             } else {
1224                 $gt = sq_findnxstr($body, $pos, ">");
1225                 $retary = Array(false, false, false, $lt, $gt);
1226                 return $retary;
1227             }
1228         case '>':
1229             return Array($tagname, false, $tagtype, $lt, $pos);
1230             break;
1231         default:
1232             /**
1233             * Check if it's whitespace
1234             */
1235             if (!preg_match('/\s/', $match)){
1236                 /**
1237                 * This is an invalid tag! Look for the next closing ">".
1238                 */
1239                 $gt = sq_findnxstr($body, $lt, ">");
1240                 return Array(false, false, false, $lt, $gt);
1241             }
1242             break;
1243     }
1244
1245     /**
1246     * At this point we're here:
1247     * <tagname  attribute='blah'>
1248     * \-------^
1249     *
1250     * At this point we loop in order to find all attributes.
1251     */
1252     $attname = '';
1253     $attary = Array();
1254
1255     while ($pos <= strlen($body)){
1256         $pos = sq_skipspace($body, $pos);
1257         if ($pos == strlen($body)){
1258             /**
1259             * Non-closed tag.
1260             */
1261             return Array(false, false, false, $lt, $pos);
1262         }
1263         /**
1264         * See if we arrived at a ">" or "/>", which means that we reached
1265         * the end of the tag.
1266         */
1267         $matches = Array();
1268         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1269             /**
1270             * Yep. So we did.
1271             */
1272             $pos += strlen($matches{1});
1273             if ($matches{2} == "/>"){
1274                 $tagtype = 3;
1275                 $pos++;
1276             }
1277             return Array($tagname, $attary, $tagtype, $lt, $pos);
1278         }
1279
1280         /**
1281         * There are several types of attributes, with optional
1282         * [:space:] between members.
1283         * Type 1:
1284         *   attrname[:space:]=[:space:]'CDATA'
1285         * Type 2:
1286         *   attrname[:space:]=[:space:]"CDATA"
1287         * Type 3:
1288         *   attr[:space:]=[:space:]CDATA
1289         * Type 4:
1290         *   attrname
1291         *
1292         * We leave types 1 and 2 the same, type 3 we check for
1293         * '"' and convert to "&quot" if needed, then wrap in
1294         * double quotes. Type 4 we convert into:
1295         * attrname="yes".
1296         */
1297         $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1298         if ($regary == false){
1299             /**
1300             * Looks like body ended before the end of tag.
1301             */
1302             return Array(false, false, false, $lt, strlen($body));
1303         }
1304         list($pos, $attname, $match) = $regary;
1305         $attname = strtolower($attname);
1306         /**
1307         * We arrived at the end of attribute name. Several things possible
1308         * here:
1309         * '>'  means the end of the tag and this is attribute type 4
1310         * '/'  if followed by '>' means the same thing as above
1311         * '\s' means a lot of things -- look what it's followed by.
1312         *      anything else means the attribute is invalid.
1313         */
1314         switch($match){
1315             case '/':
1316                 /**
1317                 * This is an xhtml-style tag with a closing / at the
1318                 * end, like so: <img src="blah" />. Check if it's followed
1319                 * by the closing bracket. If not, then this tag is invalid
1320                 */
1321                 if (substr($body, $pos, 2) == "/>"){
1322                     $pos++;
1323                     $tagtype = 3;
1324                 } else {
1325                     $gt = sq_findnxstr($body, $pos, ">");
1326                     $retary = Array(false, false, false, $lt, $gt);
1327                     return $retary;
1328                 }
1329             case '>':
1330                 $attary{$attname} = '"yes"';
1331                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1332                 break;
1333             default:
1334                 /**
1335                 * Skip whitespace and see what we arrive at.
1336                 */
1337                 $pos = sq_skipspace($body, $pos);
1338                 $char = substr($body, $pos, 1);
1339                 /**
1340                 * Two things are valid here:
1341                 * '=' means this is attribute type 1 2 or 3.
1342                 * \w means this was attribute type 4.
1343                 * anything else we ignore and re-loop. End of tag and
1344                 * invalid stuff will be caught by our checks at the beginning
1345                 * of the loop.
1346                 */
1347                 if ($char == "="){
1348                     $pos++;
1349                     $pos = sq_skipspace($body, $pos);
1350                     /**
1351                     * Here are 3 possibilities:
1352                     * "'"  attribute type 1
1353                     * '"'  attribute type 2
1354                     * everything else is the content of tag type 3
1355                     */
1356                     $quot = substr($body, $pos, 1);
1357                     if ($quot == "'"){
1358                         $regary = sq_findnxreg($body, $pos+1, "\'");
1359                         if ($regary == false){
1360                             return Array(false, false, false, $lt, strlen($body));
1361                         }
1362                         list($pos, $attval, $match) = $regary;
1363                         $pos++;
1364                         $attary{$attname} = "'" . $attval . "'";
1365                     } else if ($quot == '"'){
1366                         $regary = sq_findnxreg($body, $pos+1, '\"');
1367                         if ($regary == false){
1368                             return Array(false, false, false, $lt, strlen($body));
1369                         }
1370                         list($pos, $attval, $match) = $regary;
1371                         $pos++;
1372                         $attary{$attname} = '"' . $attval . '"';
1373                     } else {
1374                         /**
1375                         * These are hateful. Look for \s, or >.
1376                         */
1377                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1378                         if ($regary == false){
1379                             return Array(false, false, false, $lt, strlen($body));
1380                         }
1381                         list($pos, $attval, $match) = $regary;
1382                         /**
1383                         * If it's ">" it will be caught at the top.
1384                         */
1385                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1386                         $attary{$attname} = '"' . $attval . '"';
1387                     }
1388                 } else if (preg_match("|[\w/>]|", $char)) {
1389                     /**
1390                     * That was attribute type 4.
1391                     */
1392                     $attary{$attname} = '"yes"';
1393                 } else {
1394                     /**
1395                     * An illegal character. Find next '>' and return.
1396                     */
1397                     $gt = sq_findnxstr($body, $pos, ">");
1398                     return Array(false, false, false, $lt, $gt);
1399                 }
1400                 break;
1401         }
1402     }
1403     /**
1404     * The fact that we got here indicates that the tag end was never
1405     * found. Return invalid tag indication so it gets stripped.
1406     */
1407     return Array(false, false, false, $lt, strlen($body));
1408 }
1409
1410 /**
1411 * This function checks attribute values for entity-encoded values
1412 * and returns them translated into 8-bit strings so we can run
1413 * checks on them.
1414 *
1415 * @param  $attvalue A string to run entity check against.
1416 * @return           Translated value.
1417 */
1418
1419 function sq_deent($attvalue){
1420     $me = 'sq_deent';
1421     /**
1422     * See if we have to run the checks first. All entities must start
1423     * with "&".
1424     */
1425     if (strpos($attvalue, '&') === false){
1426         return $attvalue;
1427     }
1428     /**
1429     * Check named entities first.
1430     */
1431     $trans = get_html_translation_table(HTML_ENTITIES);
1432     /**
1433     * Leave &quot; in, as it can mess us up.
1434     */
1435     $trans = array_flip($trans);
1436     unset($trans{'&quot;'});
1437     while (list($ent, $val) = each($trans)){
1438         $attvalue = preg_replace('/' . $ent . '*/si', $val, $attvalue);
1439     }
1440     /**
1441     * Now translate numbered entities from 1 to 255 if needed.
1442     */
1443     if (strpos($attvalue, '#') !== false){
1444         $omit = Array(34, 39);
1445         for ($asc = 256; $asc >= 0; $asc--){
1446             if (!in_array($asc, $omit)){
1447                 $chr = chr($asc);
1448                 $octrule = '/\&#0*' . $asc . ';*/si';
1449                 $hexrule = '/\&#x0*' . dechex($asc) . ';*/si';
1450                 $attvalue = preg_replace($octrule, $chr, $attvalue);
1451                 $attvalue = preg_replace($hexrule, $chr, $attvalue);
1452             }
1453         }
1454     }
1455     return $attvalue;
1456 }
1457
1458 /**
1459 * This function runs various checks against the attributes.
1460 *
1461 * @param  $tagname         String with the name of the tag.
1462 * @param  $attary          Array with all tag attributes.
1463 * @param  $rm_attnames     See description for sq_sanitize
1464 * @param  $bad_attvals     See description for sq_sanitize
1465 * @param  $add_attr_to_tag See description for sq_sanitize
1466 * @param  $message         message object
1467 * @param  $id              message id
1468 * @return                  Array with modified attributes.
1469 */
1470 function sq_fixatts($tagname,
1471                     $attary,
1472                     $rm_attnames,
1473                     $bad_attvals,
1474                     $add_attr_to_tag,
1475                     $message,
1476                     $id,
1477                     $mailbox
1478                     ){
1479     $me = 'sq_fixatts';
1480     while (list($attname, $attvalue) = each($attary)){
1481         /**
1482         * See if this attribute should be removed.
1483         */
1484         foreach ($rm_attnames as $matchtag=>$matchattrs){
1485             if (preg_match($matchtag, $tagname)){
1486                 foreach ($matchattrs as $matchattr){
1487                     if (preg_match($matchattr, $attname)){
1488                         unset($attary{$attname});
1489                         continue;
1490                     }
1491                 }
1492             }
1493         }
1494         /**
1495         * Remove any backslashes, entities, and extraneous whitespace.
1496         */
1497         $attvalue = sq_unbackslash($attvalue);
1498         $attvalue = sq_deent($attvalue);
1499         $attvalue = sq_unspace($attvalue);
1500
1501         /**
1502         * Remove \r \n \t \0 " " "\\"
1503         */
1504         $attvalue = str_replace(Array("\r", "\n", "\t", "\0", " ", "\\"),
1505                         Array('', '','','','',''), $attvalue);
1506
1507         /**
1508         * Now let's run checks on the attvalues.
1509         * I don't expect anyone to comprehend this. If you do,
1510         * get in touch with me so I can drive to where you live and
1511         * shake your hand personally. :)
1512         */
1513         foreach ($bad_attvals as $matchtag=>$matchattrs){
1514             if (preg_match($matchtag, $tagname)){
1515                 foreach ($matchattrs as $matchattr=>$valary){
1516                     if (preg_match($matchattr, $attname)){
1517                         /**
1518                         * There are two arrays in valary.
1519                         * First is matches.
1520                         * Second one is replacements
1521                         */
1522                         list($valmatch, $valrepl) = $valary;
1523                         $newvalue =
1524                             preg_replace($valmatch, $valrepl, $attvalue);
1525                         if ($newvalue != $attvalue){
1526                             $attary{$attname} = $newvalue;
1527                         }
1528                     }
1529                 }
1530             }
1531         }
1532
1533
1534         /**
1535          * Replace empty src tags with the blank image.  src is only used
1536          * for frames, images, and image inputs.  Doing a replace should
1537          * not affect them working as should be, however it will stop
1538          * IE from being kicked off when src for img tags are not set
1539          */
1540         if (($attname == 'src') && ($attvalue == '""')) {
1541             $attary{$attname} = '"' . SM_PATH . 'images/blank.png"';
1542         }
1543
1544         /**
1545         * Turn cid: urls into http-friendly ones.
1546         */
1547         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1548             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1549         }
1550
1551         /**
1552          * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1553          * One day MS might actually make it match something useful, for now, falling
1554          * back to using cid2http, so we can grab the blank.png.
1555          */
1556         if (preg_match("/^[\'\"]\s*outbind:\/\//si", $attvalue)) {
1557             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1558         }
1559
1560     }
1561     /**
1562     * See if we need to append any attributes to this tag.
1563     */
1564     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1565         if (preg_match($matchtag, $tagname)){
1566             $attary = array_merge($attary, $addattary);
1567         }
1568     }
1569     return $attary;
1570 }
1571
1572 /**
1573 * This function edits the style definition to make them friendly and
1574 * usable in SquirrelMail.
1575 *
1576 * @param  $message  the message object
1577 * @param  $id       the message id
1578 * @param  $content  a string with whatever is between <style> and </style>
1579 * @param  $mailbox  the message mailbox
1580 * @return           a string with edited content.
1581 */
1582 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1583     global $view_unsafe_images;
1584     $me = 'sq_fixstyle';
1585     $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1586     if ($ret == FALSE){
1587         return array(FALSE, strlen($body));
1588     }
1589     $newpos = $ret[0] + strlen($ret[2]);
1590     $content = $ret[1];
1591     /**
1592     * First look for general BODY style declaration, which would be
1593     * like so:
1594     * body {background: blah-blah}
1595     * and change it to .bodyclass so we can just assign it to a <div>
1596     */
1597     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1598     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1599     /**
1600     * Fix url('blah') declarations.
1601     */
1602     $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1603                             "url(\\1$secremoveimg\\2)", $content);
1604     /**
1605     * Fix url('https*://.*) declarations but only if $view_unsafe_images
1606     * is false.
1607     */
1608     if (!$view_unsafe_images){
1609         $content = preg_replace("|url\s*\(\s*([\'\"])\s*https*:.*?([\'\"])\s*\)|si",
1610                                 "url(\\1$secremoveimg\\2)", $content);
1611     }
1612
1613     /**
1614     * Fix urls that refer to cid:
1615     */
1616     while (preg_match("|url\s*\(\s*([\'\"]\s*cid:.*?[\'\"])\s*\)|si",
1617                     $content, $matches)){
1618         $cidurl = $matches{1};
1619         $httpurl = sq_cid2http($message, $id, $cidurl, $mailbox);
1620         $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1621                                 "url($httpurl)", $content);
1622     }
1623
1624     /**
1625     * Fix stupid css declarations which lead to vulnerabilities
1626     * in IE.
1627     */
1628     $match   = Array('/expression/i',
1629                     '/behaviou*r/i',
1630                     '/binding/i',
1631                     '/include-source/i');
1632     $replace = Array('idiocy', 'idiocy', 'idiocy', 'idiocy');
1633     $content = preg_replace($match, $replace, $content);
1634     return array($content, $newpos);
1635 }
1636
1637 /**
1638 * This function converts cid: url's into the ones that can be viewed in
1639 * the browser.
1640 *
1641 * @param  $message  the message object
1642 * @param  $id       the message id
1643 * @param  $cidurl   the cid: url.
1644 * @param  $mailbox  the message mailbox
1645 * @return           a string with a http-friendly url
1646 */
1647 function sq_cid2http($message, $id, $cidurl, $mailbox){
1648     /**
1649     * Get rid of quotes.
1650     */
1651     $quotchar = substr($cidurl, 0, 1);
1652     if ($quotchar == '"' || $quotchar == "'"){
1653         $cidurl = str_replace($quotchar, "", $cidurl);
1654     } else {
1655         $quotchar = '';
1656     }
1657     $cidurl = substr(trim($cidurl), 4);
1658     $linkurl = find_ent_id($cidurl, $message);
1659     /* in case of non-save cid links $httpurl should be replaced by a sort of
1660     unsave link image */
1661     $httpurl = '';
1662
1663    /**
1664     * This is part of a fix for Outlook Express 6.x generating
1665     * cid URLs without creating content-id headers. These images are
1666     * not part of the multipart/related html mail. The html contains
1667     * <img src="cid:{some_id}/image_filename.ext"> references to
1668     * attached images with as goal to render them inline although
1669     * the attachment disposition property is not inline.
1670     **/
1671
1672     if (empty($linkurl)) {
1673         if (preg_match('/{.*}\//', $cidurl)) {
1674             $cidurl = preg_replace('/{.*}\//','', $cidurl);
1675             if (!empty($cidurl)) {
1676                 $linkurl = find_ent_id($cidurl, $message);
1677             }
1678         }
1679     }
1680
1681     if (!empty($linkurl)) {
1682         $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1683                    "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1684                    '&amp;ent_id=' . $linkurl . $quotchar;
1685     } else {
1686         /**
1687          * If we couldn't generate a proper img url, drop in a blank image
1688          * instead of sending back empty, otherwise it causes unusual behaviour
1689          */
1690         $httpurl = $quotchar . SM_PATH . 'images/blank.png';
1691     }
1692
1693     return $httpurl;
1694 }
1695
1696 /**
1697 * This function changes the <body> tag into a <div> tag since we
1698 * can't really have a body-within-body.
1699 *
1700 * @param  $attary   an array of attributes and values of <body>
1701 * @param  $mailbox  mailbox we're currently reading (for cid2http)
1702 * @param  $message  current message (for cid2http)
1703 * @param  $id       current message id (for cid2http)
1704 * @return           a modified array of attributes to be set for <div>
1705 */
1706 function sq_body2div($attary, $mailbox, $message, $id){
1707     $me = 'sq_body2div';
1708     $divattary = Array('class' => "'bodyclass'");
1709     $text = '#000000';
1710     $has_bgc_stl = $has_txt_stl = false;
1711     $styledef = '';
1712     if (is_array($attary) && sizeof($attary) > 0){
1713         foreach ($attary as $attname=>$attvalue){
1714             $quotchar = substr($attvalue, 0, 1);
1715             $attvalue = str_replace($quotchar, "", $attvalue);
1716             switch ($attname){
1717                 case 'background':
1718                     $attvalue = sq_cid2http($message, $id,
1719                                             $attvalue, $mailbox);
1720                     $styledef .= "background-image: url('$attvalue'); ";
1721                     break;
1722                 case 'bgcolor':
1723                     $has_bgc_stl = true;
1724                     $styledef .= "background-color: $attvalue; ";
1725                     break;
1726                 case 'text':
1727                     $has_txt_stl = true;
1728                     $styledef .= "color: $attvalue; ";
1729                     break;
1730             }
1731         }
1732         // Outlook defines a white bgcolor and no text color. This can lead to
1733         // white text on a white bg with certain themes.
1734         if ($has_bgc_stl && !$has_txt_stl) {
1735             $styledef .= "color: $text; ";
1736         }
1737         if (strlen($styledef) > 0){
1738             $divattary{"style"} = "\"$styledef\"";
1739         }
1740     }
1741     return $divattary;
1742 }
1743
1744 /**
1745 * This is the main function and the one you should actually be calling.
1746 * There are several variables you should be aware of an which need
1747 * special description.
1748 *
1749 * Since the description is quite lengthy, see it here:
1750 * http://linux.duke.edu/projects/mini/htmlfilter/
1751 *
1752 * @param $body                 the string with HTML you wish to filter
1753 * @param $tag_list             see description above
1754 * @param $rm_tags_with_content see description above
1755 * @param $self_closing_tags    see description above
1756 * @param $force_tag_closing    see description above
1757 * @param $rm_attnames          see description above
1758 * @param $bad_attvals          see description above
1759 * @param $add_attr_to_tag      see description above
1760 * @param $message              message object
1761 * @param $id                   message id
1762 * @return                      sanitized html safe to show on your pages.
1763 */
1764 function sq_sanitize($body,
1765                     $tag_list,
1766                     $rm_tags_with_content,
1767                     $self_closing_tags,
1768                     $force_tag_closing,
1769                     $rm_attnames,
1770                     $bad_attvals,
1771                     $add_attr_to_tag,
1772                     $message,
1773                     $id,
1774                     $mailbox
1775                     ){
1776     $me = 'sq_sanitize';
1777     $rm_tags = array_shift($tag_list);
1778     /**
1779     * Normalize rm_tags and rm_tags_with_content.
1780     */
1781     @array_walk($tag_list, 'sq_casenormalize');
1782     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1783     @array_walk($self_closing_tags, 'sq_casenormalize');
1784     /**
1785     * See if tag_list is of tags to remove or tags to allow.
1786     * false  means remove these tags
1787     * true   means allow these tags
1788     */
1789     $curpos = 0;
1790     $open_tags = Array();
1791     $trusted = "\n<!-- begin sanitized html -->\n";
1792     $skip_content = false;
1793     /**
1794     * Take care of netscape's stupid javascript entities like
1795     * &{alert('boo')};
1796     */
1797     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1798
1799     while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1800         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1801         $free_content = substr($body, $curpos, $lt-$curpos);
1802         /**
1803         * Take care of <style>
1804         */
1805         if ($tagname == "style" && $tagtype == 1){
1806             list($free_content, $curpos) =
1807                 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
1808             if ($free_content != FALSE){
1809                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1810                 $trusted .= $free_content;
1811                 $trusted .= sq_tagprint($tagname, false, 2);
1812             }
1813             continue;
1814         }
1815         if ($skip_content == false){
1816             $trusted .= $free_content;
1817         }
1818         if ($tagname != FALSE){
1819             if ($tagtype == 2){
1820                 if ($skip_content == $tagname){
1821                     /**
1822                     * Got to the end of tag we needed to remove.
1823                     */
1824                     $tagname = false;
1825                     $skip_content = false;
1826                 } else {
1827                     if ($skip_content == false){
1828                         if ($tagname == "body"){
1829                             $tagname = "div";
1830                         }
1831                         if (isset($open_tags{$tagname}) &&
1832                             $open_tags{$tagname} > 0){
1833                             $open_tags{$tagname}--;
1834                         } else {
1835                             $tagname = false;
1836                         }
1837                     }
1838                 }
1839             } else {
1840                 /**
1841                 * $rm_tags_with_content
1842                 */
1843                 if ($skip_content == false){
1844                     /**
1845                     * See if this is a self-closing type and change
1846                     * tagtype appropriately.
1847                     */
1848                     if ($tagtype == 1
1849                         && in_array($tagname, $self_closing_tags)){
1850                         $tagtype = 3;
1851                     }
1852                     /**
1853                     * See if we should skip this tag and any content
1854                     * inside it.
1855                     */
1856                     if ($tagtype == 1 &&
1857                         in_array($tagname, $rm_tags_with_content)){
1858                         $skip_content = $tagname;
1859                     } else {
1860                         if (($rm_tags == false
1861                             && in_array($tagname, $tag_list)) ||
1862                             ($rm_tags == true &&
1863                             !in_array($tagname, $tag_list))){
1864                             $tagname = false;
1865                         } else {
1866                             /**
1867                             * Convert body into div.
1868                             */
1869                             if ($tagname == "body"){
1870                                 $tagname = "div";
1871                                 $attary = sq_body2div($attary, $mailbox,
1872                                                     $message, $id);
1873                             }
1874                             if ($tagtype == 1){
1875                                 if (isset($open_tags{$tagname})){
1876                                     $open_tags{$tagname}++;
1877                                 } else {
1878                                     $open_tags{$tagname}=1;
1879                                 }
1880                             }
1881                             /**
1882                             * This is where we run other checks.
1883                             */
1884                             if (is_array($attary) && sizeof($attary) > 0){
1885                                 $attary = sq_fixatts($tagname,
1886                                                     $attary,
1887                                                     $rm_attnames,
1888                                                     $bad_attvals,
1889                                                     $add_attr_to_tag,
1890                                                     $message,
1891                                                     $id,
1892                                                     $mailbox
1893                                                     );
1894                             }
1895                         }
1896                     }
1897                 }
1898             }
1899             if ($tagname != false && $skip_content == false){
1900                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1901             }
1902         }
1903         $curpos = $gt+1;
1904     }
1905     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1906     if ($force_tag_closing == true){
1907         foreach ($open_tags as $tagname=>$opentimes){
1908             while ($opentimes > 0){
1909                 $trusted .= '</' . $tagname . '>';
1910                 $opentimes--;
1911             }
1912         }
1913         $trusted .= "\n";
1914     }
1915     $trusted .= "<!-- end sanitized html -->\n";
1916     return $trusted;
1917 }
1918
1919 /**
1920 * This is a wrapper function to call html sanitizing routines.
1921 *
1922 * @param  $body  the body of the message
1923 * @param  $id    the id of the message
1924 * @param  $message
1925 * @param  $mailbox
1926 * @param  boolean $take_mailto_links When TRUE, converts mailto: links
1927 *                                    into internal SM compose links
1928 *                                    (optional; default = TRUE)
1929 * @return        a string with html safe to display in the browser.
1930 */
1931 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links = true) {
1932
1933     require_once(SM_PATH . 'functions/url_parser.php');  // for $MailTo_PReg_Match
1934
1935     global $attachment_common_show_images, $view_unsafe_images,
1936         $has_unsafe_images;
1937
1938     /**
1939     * Don't display attached images in HTML mode.
1940     */
1941     $attachment_common_show_images = false;
1942     $tag_list = Array(
1943                     false,
1944                     "object",
1945                     "meta",
1946                     "html",
1947                     "head",
1948                     "base",
1949                     "link",
1950                     "frame",
1951                     "iframe",
1952                     "plaintext",
1953                     "marquee"
1954                     );
1955
1956     $rm_tags_with_content = Array(
1957                                 "script",
1958                                 "applet",
1959                                 "embed",
1960                                 "title",
1961                                 "frameset",
1962                                 "xml"
1963                                 );
1964
1965     $self_closing_tags =  Array(
1966                                 "img",
1967                                 "br",
1968                                 "hr",
1969                                 "input",
1970                                 "outbind"
1971                                 );
1972
1973     $force_tag_closing = true;
1974
1975     $rm_attnames = Array(
1976                         "/.*/" =>
1977                         Array(
1978                             "/target/i",
1979                             "/^on.*/i",
1980                             "/^dynsrc/i",
1981                             "/^data.*/i",
1982                             "/^lowsrc.*/i"
1983                             )
1984                         );
1985
1986     $secremoveimg = "../images/" . _("sec_remove_eng.png");
1987     $bad_attvals = Array(
1988         "/.*/" =>
1989             Array(
1990                 "/^src|background/i" =>
1991                     Array(
1992                         Array(
1993                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1994                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1995                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1996                                 ),
1997                         Array(
1998                                 "\\1$secremoveimg\\2",
1999                                 "\\1$secremoveimg\\2",
2000                                 "\\1$secremoveimg\\2",
2001                                 "\\1$secremoveimg\\2"
2002                                 )
2003                         ),
2004                 "/^href|action/i" =>
2005                     Array(
2006                         Array(
2007                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2008                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2009                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2010                                 ),
2011                         Array(
2012                                 "\\1#\\1",
2013                                 "\\1#\\1",
2014                                 "\\1#\\1",
2015                                 "\\1#\\1"
2016                                 )
2017                         ),
2018                 "/^style/i" =>
2019                     Array(
2020                         Array(
2021                                 "/expression/i",
2022                                 "/binding/i",
2023                                 "/behaviou*r/i",
2024                                 "/include-source/i",
2025                                 "/position\s*:\s*absolute/i",
2026                                 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2027                                 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2028                                 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2029                                 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2030                             ),
2031                         Array(
2032                                 "idiocy",
2033                                 "idiocy",
2034                                 "idiocy",
2035                                 "idiocy",
2036                                 "",
2037                                 "url(\\1#\\1)",
2038                                 "url(\\1#\\1)",
2039                                 "url(\\1#\\1)",
2040                                 "url(\\1#\\1)",
2041                                 "\\1:url(\\2#\\3)"
2042                             )
2043                         )
2044                 )
2045         );
2046     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
2047         $view_unsafe_images = false;
2048     }
2049     if (!$view_unsafe_images){
2050         /**
2051         * Remove any references to http/https if view_unsafe_images set
2052         * to false.
2053         */
2054         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2055                     '/^([\'\"])\s*https*:.*([\'\"])/si');
2056         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2057                     "\\1$secremoveimg\\1");
2058         array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2059                     '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
2060         array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2061                     "url(\\1$secremoveimg\\1)");
2062     }
2063
2064     $add_attr_to_tag = Array(
2065         "/^a$/i" =>
2066             Array('target'=>'"_blank"',
2067                 'title'=>'"'._("This external link will open in a new window").'"'
2068             )
2069     );
2070     $trusted = sq_sanitize($body,
2071                         $tag_list,
2072                         $rm_tags_with_content,
2073                         $self_closing_tags,
2074                         $force_tag_closing,
2075                         $rm_attnames,
2076                         $bad_attvals,
2077                         $add_attr_to_tag,
2078                         $message,
2079                         $id,
2080                         $mailbox
2081                         );
2082     if (preg_match("|$secremoveimg|i", $trusted)){
2083         $has_unsafe_images = true;
2084     }
2085
2086
2087     // we want to parse mailto's in HTML output, change to SM compose links
2088     // this is a modified version of code from url_parser.php... but Marc is
2089     // right: we need a better filtering implementation; adding this randomly
2090     // here is not a great solution
2091     //
2092     if ($take_mailto_links) {
2093         // parseUrl($trusted);   // this even parses URLs inside of tags... too aggressive
2094         global $MailTo_PReg_Match;
2095         $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1);
2096         if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2097             foreach ($regs[0] as $i => $mailto_before) {
2098                 $mailto_params = $regs[10][$i];
2099
2100                 // get rid of any tailing quote since we have to add send_to to the end
2101                 //
2102                 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2103                     $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2104                 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2105                     $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2106
2107                 if ($regs[1][$i]) {    //if there is an email addr before '?', we need to merge it with the params
2108                     $to = 'to=' . $regs[1][$i];
2109                     if (strpos($mailto_params, 'to=') > -1)    //already a 'to='
2110                         $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2111                     else {
2112                         if ($mailto_params)    //already some params, append to them
2113                             $mailto_params .= '&amp;' . $to;
2114                         else
2115                             $mailto_params .= '?' . $to;
2116                     }
2117                 }
2118
2119                 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2120
2121                 // we'll already have target=_blank, no need to allow comp_in_new
2122                 // here (which would be a lot more work anyway)
2123                 //
2124                 global $compose_new_win;
2125                 $temp_comp_in_new = $compose_new_win;
2126                 $compose_new_win = 0;
2127                 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2128                 $compose_new_win = $temp_comp_in_new;
2129
2130                 // remove <a href=" and anything after the next quote (we only
2131                 // need the uri, not the link HTML) in compose uri
2132                 //
2133                 $comp_uri = substr($comp_uri, 9);
2134                 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2135                 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2136             }
2137         }
2138     }
2139
2140     return $trusted;
2141 }
2142
2143 /**
2144 * function SendDownloadHeaders - send file to the browser
2145 *
2146 * Original Source: SM core src/download.php
2147 * moved here to make it available to other code, and separate
2148 * front end from back end functionality.
2149 *
2150 * @param string $type0 first half of mime type
2151 * @param string $type1 second half of mime type
2152 * @param string $filename filename to tell the browser for downloaded file
2153 * @param boolean $force whether to force the download dialog to pop
2154 * @param optional integer $filesize send the Content-Header and length to the browser
2155 * @return void
2156 */
2157 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2158     global $languages, $squirrelmail_language;
2159     $isIE = $isIE6 = 0;
2160
2161     sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2162
2163     if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2164         strstr($HTTP_USER_AGENT, 'Opera') === false) {
2165         $isIE = 1;
2166     }
2167
2168     if (strstr($HTTP_USER_AGENT, 'compatible; MSIE 6') !== false &&
2169         strstr($HTTP_USER_AGENT, 'Opera') === false) {
2170         $isIE6 = 1;
2171     }
2172
2173     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2174         function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2175         $filename =
2176         call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2177     } else {
2178         $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
2179     }
2180
2181     // A Pox on Microsoft and it's Internet Explorer!
2182     //
2183     // IE has lots of bugs with file downloads.
2184     // It also has problems with SSL.  Both of these cause problems
2185     // for us in this function.
2186     //
2187     // See this article on Cache Control headers and SSL
2188     // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2189     //
2190     // The best thing you can do for IE is to upgrade to the latest
2191     // version
2192     //set all the Cache Control Headers for IE
2193     if ($isIE) {
2194         $filename=rawurlencode($filename);
2195         header ("Pragma: public");
2196         header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); # HTTP/1.1
2197         header ("Cache-Control: post-check=0, pre-check=0", false);
2198         header ("Cache-control: private");
2199
2200         //set the inline header for IE, we'll add the attachment header later if we need it
2201         header ("Content-Disposition: inline; filename=$filename");
2202     }
2203
2204     if (!$force) {
2205         // Try to show in browser window
2206         header ("Content-Disposition: inline; filename=\"$filename\"");
2207         header ("Content-Type: $type0/$type1; name=\"$filename\"");
2208     } else {
2209         // Try to pop up the "save as" box
2210
2211         // IE makes this hard.  It pops up 2 save boxes, or none.
2212         // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2213         // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2214         // But, according to Microsoft, it is "RFC compliant but doesn't
2215         // take into account some deviations that allowed within the
2216         // specification."  Doesn't that mean RFC non-compliant?
2217         // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2218
2219         // all browsers need the application/octet-stream header for this
2220         header ("Content-Type: application/octet-stream; name=\"$filename\"");
2221
2222         // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2223         // Do not have quotes around filename, but that applied to
2224         // "attachment"... does it apply to inline too?
2225         header ("Content-Disposition: attachment; filename=\"$filename\"");
2226
2227         if ($isIE && !$isIE6) {
2228             // This combination seems to work mostly.  IE 5.5 SP 1 has
2229             // known issues (see the Microsoft Knowledge Base)
2230
2231             // This works for most types, but doesn't work with Word files
2232             header ("Content-Type: application/download; name=\"$filename\"");
2233
2234             // These are spares, just in case.  :-)
2235             //header("Content-Type: $type0/$type1; name=\"$filename\"");
2236             //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2237             //header("Content-Type: application/octet-stream; name=\"$filename\"");
2238         } else {
2239             // another application/octet-stream forces download for Netscape
2240             header ("Content-Type: application/octet-stream; name=\"$filename\"");
2241         }
2242     }
2243
2244     //send the content-length header if the calling function provides it
2245     if ($filesize > 0) {
2246         header("Content-Length: $filesize");
2247     }
2248
2249 }  // end fn SendDownloadHeaders
2250
2251 ?>