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 bodystructure provided 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, $rStream='php://stdout') {
 163
 164     /* Don't kill the connection if the browser is over a dialup
 165      * and it would take over 30 seconds to download it.
 166      * Don't call set_time_limit in safe mode.
 167      */
 168
 169     if (!ini_get('safe_mode')) {
 170         set_time_limit(0);
 171     }
 172     /* in case of base64 encoded attachments, do not buffer them.
 173        Instead, echo the decoded attachment directly to screen */
 174     if (strtolower($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',$rStream,true);
 181     } else {
 182         $body = mime_fetch_body ($imap_stream, $id, $ent_id);
 183         if (is_resource($rStream)) {
 184             fputs($rStream,decodeBody($body,$encoding));
 185         } else {
 186             echo decodeBody($body, $encoding);
 187         }
 188     }
 189
 190     /*
 191        TODO, use the same method for quoted printable.
 192        However, I assume that quoted printable attachments aren't that large
 193        so the performancegain / memory usage drop will be minimal.
 194        If we decide to add that then we need to adapt sqimap_fread because
 195        we need to split te result on \n and fread doesn't stop at \n. That
 196        means we also should provide $results from sqimap_fread (by ref) to
 197        te function and set $no_return to false. The $filter function for
 198        quoted printable should handle unsetting of $results.
 199      */
 200     /*
 201        TODO 2: find out how we write to the output stream php://stdout. fwrite
 202        doesn't work because 'php://stdout isn't a stream.
 203      */
 204
 205     return;
 206 }
 207
 208 /* -[ END MIME DECODING ]----------------------------------------------------------- */
 209
 210 /* This is here for debugging purposes.  It will print out a list
 211  * of all the entity IDs that are in the $message object.
 212  */
 213 function listEntities ($message) {
 214     if ($message) {
 215         echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br />';
 216         for ($i = 0; isset($message->entities[$i]); $i++) {
 217             echo "$i : ";
 218             $msg = listEntities($message->entities[$i]);
 219
 220             if ($msg) {
 221                 echo "return: ";
 222                 return $msg;
 223             }
 224         }
 225     }
 226 }
 227
 228 function getPriorityStr($priority) {
 229     $priority_level = substr($priority,0,1);
 230
 231     switch($priority_level) {
 232         /* Check for a higher then normal priority. */
 233         case '1':
 234         case '2':
 235             $priority_string = _("High");
 236             break;
 237
 238         /* Check for a lower then normal priority. */
 239         case '4':
 240         case '5':
 241             $priority_string = _("Low");
 242             break;
 243
 244         /* Check for a normal priority. */
 245         case '3':
 246         default:
 247             $priority_level = '3';
 248             $priority_string = _("Normal");
 249             break;
 250
 251     }
 252     return $priority_string;
 253 }
 254
 255 /* returns a $message object for a particular entity id */
 256 function getEntity ($message, $ent_id) {
 257     return $message->getEntity($ent_id);
 258 }
 259
 260 /* translateText
 261  * Extracted from strings.php 23/03/2002
 262  */
 263
 264 function translateText(&$body, $wrap_at, $charset) {
 265     global $where, $what;   /* from searching */
 266     global $color;          /* color theme */
 267
 268     require_once(SM_PATH . 'functions/url_parser.php');
 269
 270     $body_ary = explode("\n", $body);
 271     for ($i=0; $i < count($body_ary); $i++) {
 272         $line = $body_ary[$i];
 273         if (strlen($line) - 2 >= $wrap_at) {
 274             sqWordWrap($line, $wrap_at, $charset);
 275         }
 276         $line = charset_decode($charset, $line);
 277         $line = str_replace("\t", '        ', $line);
 278
 279         parseUrl ($line);
 280
 281         $quotes = 0;
 282         $pos = 0;
 283         $j = strlen($line);
 284
 285         while ($pos < $j) {
 286             if ($line[$pos] == ' ') {
 287                 $pos++;
 288             } else if (strpos($line, '&gt;', $pos) === $pos) {
 289                 $pos += 4;
 290                 $quotes++;
 291             } else {
 292                 break;
 293             }
 294         }
 295
 296         if ($quotes % 2) {
 297             if (!isset($color[13])) {
 298                 $color[13] = '#800000';
 299             }
 300             $line = '<font color="' . $color[13] . '">' . $line . '</font>';
 301         } elseif ($quotes) {
 302             if (!isset($color[14])) {
 303                 $color[14] = '#FF0000';
 304             }
 305             $line = '<font color="' . $color[14] . '">' . $line . '</font>';
 306         }
 307
 308         $body_ary[$i] = $line;
 309     }
 310     $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
 311 }
 312
 313 /**
 314  * This returns a parsed string called $body. That string can then
 315  * be displayed as the actual message in the HTML. It contains
 316  * everything needed, including HTML Tags, Attachments at the
 317  * bottom, etc.
 318  *
 319  * Since 1.2.0 function uses message_body hook.
 320  * Till 1.3.0 function included output of formatAttachments().
 321  *
 322  * @param resource $imap_stream imap connection resource
 323  * @param object $message squirrelmail message object
 324  * @param array $color squirrelmail color theme array
 325  * @param integer $wrap_at number of characters per line
 326  * @param string $ent_num (since 1.3.0) message part id
 327  * @param integer $id (since 1.3.0) message id
 328  * @param string $mailbox (since 1.3.0) imap folder name
 329  * @param boolean $clean (since 1.5.1) Do not output stuff that's irrelevant for the printable version.
 330  * @return string html formated message text
 331  */
 332 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX', $clean=FALSE) {
 333     /* This if statement checks for the entity to show as the
 334      * primary message. To add more of them, just put them in the
 335      * order that is their priority.
 336      */
 337     global $startMessage, $languages, $squirrelmail_language,
 338            $show_html_default, $sort, $has_unsafe_images, $passed_ent_id, $use_iframe,$iframe_height;
 339
 340     // workaround for not updated config.php
 341     if (! isset($use_iframe)) $use_iframe = false;
 342
 343     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
 344         $view_unsafe_images = false;
 345     }
 346
 347     $body = '';
 348     $urlmailbox = urlencode($mailbox);
 349     $body_message = getEntity($message, $ent_num);
 350     if (($body_message->header->type0 == 'text') ||
 351             ($body_message->header->type0 == 'rfc822')) {
 352         $body = mime_fetch_body ($imap_stream, $id, $ent_num);
 353         $body = decodeBody($body, $body_message->header->encoding);
 354
 355         if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 356                 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
 357             if (mb_detect_encoding($body) != 'ASCII') {
 358                 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
 359             }
 360         }
 361         $hookResults = do_hook("message_body", $body);
 362         $body = $hookResults[1];
 363
 364         /* If there are other types that shouldn't be formatted, add
 365          * them here.
 366          */
 367
 368         if ($body_message->header->type1 == 'html') {
 369             if ($show_html_default <> 1) {
 370                 $entity_conv = array('&nbsp;' => ' ',
 371                                      '<p>'    => "\n",
 372                                      '<P>'    => "\n",
 373                                      '<br>'   => "\n",
 374                                      '<BR>'   => "\n",
 375                                      '<br />' => "\n",
 376                                      '<BR />' => "\n",
 377                                      '&gt;'   => '>',
 378                                      '&lt;'   => '<');
 379                 $body = strtr($body, $entity_conv);
 380                 $body = strip_tags($body);
 381                 $body = trim($body);
 382                 translateText($body, $wrap_at,
 383                         $body_message->header->getParameter('charset'));
 384             } elseif ($use_iframe && ! $clean) {
 385                 // $clean is used to remove iframe in printable view.
 386
 387                 /**
 388                  * If we don't add html message between iframe tags,
 389                  * we must detect unsafe images and modify $has_unsafe_images.
 390                  */
 391                 $html_body = magicHTML($body, $id, $message, $mailbox);
 392                 // Convert character set in order to display html mails in different character set
 393                 $html_body = charset_decode($body_message->header->getParameter('charset'),$html_body,false,true);
 394
 395                 // creating iframe url
 396                 $iframeurl=sqm_baseuri().'src/view_html.php?'
 397                     . 'mailbox=' . $urlmailbox
 398                     . '&amp;passed_id=' . $id
 399                     . '&amp;ent_id=' . $ent_num
 400                     . '&amp;view_unsafe_images=' . (int) $view_unsafe_images;
 401
 402                 // adding warning message
 403                 $body = html_tag('div',_("Viewing HTML formatted email"),'center');
 404
 405                 /**
 406                  * height can't be set to 100%, because it does not work as expected when
 407                  * iframe is inside the table. Browsers do not create full height objects
 408                  * even when iframe is not nested. Maybe there is some way to get full size
 409                  * with CSS. Tested in firefox 1.02 and opera 7.53
 410                  *
 411                  * width="100%" does not work as expected, when table width is not set (automatic)
 412                  *
 413                  * tokul: I think <iframe> are safer sandbox than <object>. Objects might
 414                  * need special handling for IE and IE6SP2.
 415                  */
 416                 $body.= "<div><iframe name=\"message_frame\" width=\"100%\" height=\"$iframe_height\" src=\"$iframeurl\""
 417                     .' frameborder="1" marginwidth="0" marginheight="0" scrolling="auto">' . "\n";
 418
 419                 // Message for browsers without iframe support
 420                 //$body.= _("Your browser does not support inline frames.
 421                 // You can view HTML formated message by following below link.");
 422                 //$body.= "<br /><a href=\"$iframeurl\">"._("View HTML Message")."</a>";
 423
 424                 // if browser can't render iframe, it renders html message.
 425                 $body.= $html_body;
 426
 427                 // close iframe
 428                 $body.="</iframe></div>\n";
 429             } else {
 430                 // old way of html rendering
 431                 $body = magicHTML($body, $id, $message, $mailbox);
 432                 /**
 433                  * convert character set. charset_decode does not remove html special chars
 434                  * applied by magicHTML functions and does not sanitize them second time if
 435                  * fourth argument is true.
 436                  */
 437                 $body = charset_decode($body_message->header->getParameter('charset'),$body,false,true);
 438             }
 439         } else {
 440             translateText($body, $wrap_at,
 441                     $body_message->header->getParameter('charset'));
 442         }
 443
 444         // if this is the clean display (i.e. printer friendly), stop here.
 445         if ( $clean ) {
 446             return $body;
 447         }
 448
 449         $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
 450             '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
 451             '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 452         if (isset($passed_ent_id)) {
 453             $link .= '&amp;passed_ent_id='.$passed_ent_id;
 454         }
 455         $body .= '<center><small><a href="download.php?absolute_dl=true&amp;' .
 456             $link . '">' . _("Download this as a file") .  '</a>';
 457         if ($view_unsafe_images) {
 458             $text = _("Hide Unsafe Images");
 459         } else {
 460             if (isset($has_unsafe_images) && $has_unsafe_images) {
 461                 $link .= '&amp;view_unsafe_images=1';
 462                 $text = _("View Unsafe Images");
 463             } else {
 464                 $text = '';
 465             }
 466         }
 467         if($text != '') {
 468             $body .= '&nbsp;|&nbsp;<a href="read_body.php?' . $link . '">' . $text . '</a>';
 469         }
 470         $body .= '</small></center><br />' . "\n";
 471     }
 472     return $body;
 473 }
 474
 475 /**
 476  * Displays attachment links and information
 477  *
 478  * Since 1.3.0 function is not included in formatBody() call.
 479  *
 480  * Since 1.0.2 uses attachment $type0/$type1 hook.
 481  * Since 1.2.5 uses attachment $type0/* hook.
 482  * Since 1.5.0 uses attachments_bottom hook.
 483  *
 484  * @param object $message SquirrelMail message object
 485  * @param array $exclude_id message parts that are not attachments.
 486  * @param string $mailbox mailbox name
 487  * @param integer $id message id
 488  * @return string html formated attachment information.
 489  */
 490 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 491     global $where, $what, $startMessage, $color, $passed_ent_id;
 492
 493     $att_ar = $message->getAttachments($exclude_id);
 494
 495     if (!count($att_ar)) return '';
 496
 497     $attachments = '';
 498
 499     $urlMailbox = urlencode($mailbox);
 500
 501     foreach ($att_ar as $att) {
 502         $ent = $att->entity_id;
 503         $header = $att->header;
 504         $type0 = strtolower($header->type0);
 505         $type1 = strtolower($header->type1);
 506         $name = '';
 507         $links['download link']['text'] = _("Download");
 508         $links['download link']['href'] = sqm_baseuri() .
 509             "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 510         if ($type0 =='message' && $type1 == 'rfc822') {
 511             $default_page = sqm_baseuri() . 'src/read_body.php';
 512             $rfc822_header = $att->rfc822_header;
 513             $filename = $rfc822_header->subject;
 514             if (trim( $filename ) == '') {
 515                 $filename = 'untitled-[' . $ent . ']' ;
 516             }
 517             $from_o = $rfc822_header->from;
 518             if (is_object($from_o)) {
 519                 $from_name = decodeHeader($from_o->getAddress(false));
 520             } else {
 521                 $from_name = _("Unknown sender");
 522             }
 523             $description = $from_name;
 524         } else {
 525             $default_page = sqm_baseuri() . 'src/download.php';
 526             $filename = $att->getFilename();
 527             if ($header->description) {
 528                 $description = decodeHeader($header->description);
 529             } else {
 530                 $description = '';
 531             }
 532         }
 533
 534         $display_filename = $filename;
 535         if (isset($passed_ent_id)) {
 536             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 537         } else {
 538             $passed_ent_id_link = '';
 539         }
 540         $defaultlink = $default_page . "?startMessage=$startMessage"
 541             . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 542             . '&amp;ent_id='.$ent.$passed_ent_id_link;
 543         if ($where && $what) {
 544             $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 545         }
 546
 547         /* This executes the attachment hook with a specific MIME-type.
 548          * If that doesn't have results, it tries if there's a rule
 549          * for a more generic type.
 550          */
 551         $hookresults = do_hook("attachment $type0/$type1", $links,
 552                 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 553                 $display_filename, $where, $what);
 554         if(count($hookresults[1]) <= 1) {
 555             $hookresults = do_hook("attachment $type0/*", $links,
 556                     $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 557                     $display_filename, $where, $what);
 558         }
 559
 560         $links = $hookresults[1];
 561         $defaultlink = $hookresults[6];
 562
 563         $attachments .= '<tr><td>' .
 564             '<a href="'.$defaultlink.'">'.decodeHeader($display_filename).'</a>&nbsp;</td>' .
 565             '<td><small><b>' . show_readable_size($header->size) .
 566             '</b>&nbsp;&nbsp;</small></td>' .
 567             '<td><small>[ '.htmlspecialchars($type0).'/'.htmlspecialchars($type1).' ]&nbsp;</small></td>' .
 568             '<td><small>';
 569         $attachments .= '<b>' . $description . '</b>';
 570         $attachments .= '</small></td><td><small>&nbsp;';
 571
 572         $skipspaces = 1;
 573         foreach ($links as $val) {
 574             if ($skipspaces) {
 575                 $skipspaces = 0;
 576             } else {
 577                 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
 578             }
 579             $attachments .= '<a href="' . $val['href'] . '">'
 580                 . (isset($val['text']) && !empty($val['text']) ? $val['text'] : '')
 581                 . (isset($val['extra']) && !empty($val['extra']) ? $val['extra'] : '')
 582                 . '</a>';
 583         }
 584         unset($links);
 585         $attachments .= "</td></tr>\n";
 586     }
 587     $attachmentadd = do_hook_function('attachments_bottom',$attachments);
 588     if ($attachmentadd != '')
 589         $attachments = $attachmentadd;
 590     return $attachments;
 591 }
 592
 593 function sqimap_base64_decode(&$string) {
 594
 595     // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
 596     // fly decoding (to reduce memory usage) you have to check if the
 597     // data has incomplete pairs
 598
 599     // Remove the noise in order to check if the 4 bytes pairs are complete
 600     $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
 601
 602     $sStringRem = '';
 603     $iMod = strlen($string) % 4;
 604     if ($iMod) {
 605         $sStringRem = substr($string,-$iMod);
 606         // Check if $sStringRem contains padding characters
 607         if (substr($sStringRem,-1) != '=') {
 608             $string = substr($string,0,-$iMod);
 609         } else {
 610             $sStringRem = '';
 611         }
 612     }
 613     $string = base64_decode($string);
 614     return $sStringRem;
 615 }
 616
 617
 618 /* This function decodes the body depending on the encoding type. */
 619 function decodeBody($body, $encoding) {
 620     global $show_html_default;
 621
 622     $body = str_replace("\r\n", "\n", $body);
 623     $encoding = strtolower($encoding);
 624
 625     $encoding_handler = do_hook_function('decode_body', $encoding);
 626
 627
 628     // plugins get first shot at decoding the body
 629     //
 630     if (!empty($encoding_handler) && function_exists($encoding_handler)) {
 631         $body = $encoding_handler('decode', $body);
 632
 633     } else if ($encoding == 'quoted-printable' ||
 634             $encoding == 'quoted_printable') {
 635         $body = quoted_printable_decode($body);
 636
 637         while (ereg("=\n", $body)) {
 638             $body = ereg_replace ("=\n", '', $body);
 639         }
 640
 641     } else if ($encoding == 'base64') {
 642         $body = base64_decode($body);
 643     }
 644
 645     // All other encodings are returned raw.
 646     return $body;
 647 }
 648
 649 /**
 650  * Decodes headers
 651  *
 652  * This functions decode strings that is encoded according to
 653  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 654  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 655  *
 656  * @param string $string header string that has to be made readable
 657  * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
 658  * @param boolean $htmlsave preserve spaces and sanitize html special characters. defaults to true
 659  * @param boolean $decide decide if string can be utfencoded. defaults to false
 660  * @return string decoded header string
 661  */
 662 function decodeHeader ($string, $utfencode=true,$htmlsave=true,$decide=false) {
 663     global $languages, $squirrelmail_language,$default_charset;
 664     if (is_array($string)) {
 665         $string = implode("\n", $string);
 666     }
 667
 668     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 669             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
 670         $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
 671         // Do we need to return at this point?
 672         // return $string;
 673     }
 674     $i = 0;
 675     $iLastMatch = -2;
 676     $encoded = true;
 677
 678     $aString = explode(' ',$string);
 679     $ret = '';
 680     foreach ($aString as $chunk) {
 681         if ($encoded && $chunk === '') {
 682             continue;
 683         } elseif ($chunk === '') {
 684             $ret .= ' ';
 685             continue;
 686         }
 687         $encoded = false;
 688         /* if encoded words are not separated by a linear-space-white we still catch them */
 689         $j = $i-1;
 690
 691         while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 692             /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
 693             if ($iLastMatch !== $j) {
 694                 if ($htmlsave) {
 695                     $ret .= '&#32;';
 696                 } else {
 697                     $ret .= ' ';
 698                 }
 699             }
 700             $iLastMatch = $i;
 701             $j = $i;
 702             if ($htmlsave) {
 703                 $ret .= htmlspecialchars($res[1]);
 704             } else {
 705                 $ret .= $res[1];
 706             }
 707             $encoding = ucfirst($res[3]);
 708
 709             /* decide about valid decoding */
 710             if ($decide && is_conversion_safe($res[2])) {
 711                 $utfencode=true;
 712                 $can_be_encoded=true;
 713             } else {
 714                 $can_be_encoded=false;
 715             }
 716             switch ($encoding)
 717             {
 718                 case 'B':
 719                     $replace = base64_decode($res[4]);
 720                     if ($utfencode) {
 721                         if ($can_be_encoded) {
 722                             /* convert string to different charset,
 723                              * if functions asks for it (usually in compose)
 724                              */
 725                             $ret .= charset_convert($res[2],$replace,$default_charset);
 726                         } else {
 727                             // convert string to html codes in order to display it
 728                             $ret .= charset_decode($res[2],$replace);
 729                         }
 730                     } else {
 731                         if ($htmlsave) {
 732                             $replace = htmlspecialchars($replace);
 733                         }
 734                         $ret.= $replace;
 735                     }
 736                     break;
 737                 case 'Q':
 738                     $replace = str_replace('_', ' ', $res[4]);
 739                     $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 740                             $replace);
 741                     if ($utfencode) {
 742                         if ($can_be_encoded) {
 743                             /* convert string to different charset,
 744                              * if functions asks for it (usually in compose)
 745                              */
 746                             $replace = charset_convert($res[2], $replace,$default_charset);
 747                         } else {
 748                             // convert string to html codes in order to display it
 749                             $replace = charset_decode($res[2], $replace);
 750                         }
 751                     } else {
 752                         if ($htmlsave) {
 753                             $replace = htmlspecialchars($replace);
 754                         }
 755                     }
 756                     $ret .= $replace;
 757                     break;
 758                 default:
 759                     break;
 760             }
 761             $chunk = $res[5];
 762             $encoded = true;
 763         }
 764         if (!$encoded) {
 765             if ($htmlsave) {
 766                 $ret .= '&#32;';
 767             } else {
 768                 $ret .= ' ';
 769             }
 770         }
 771
 772         if (!$encoded && $htmlsave) {
 773             $ret .= htmlspecialchars($chunk);
 774         } else {
 775             $ret .= $chunk;
 776         }
 777         ++$i;
 778     }
 779     /* remove the first added space */
 780     if ($ret) {
 781         if ($htmlsave) {
 782             $ret = substr($ret,5);
 783         } else {
 784             $ret = substr($ret,1);
 785         }
 786     }
 787
 788     return $ret;
 789 }
 790
 791 /**
 792  * Encodes header
 793  *
 794  * Function uses XTRA_CODE _encodeheader function, if such function exists.
 795  *
 796  * Function uses Q encoding by default and encodes a string according to RFC
 797  * 1522 for use in headers if it contains 8-bit characters or anything that
 798  * looks like it should be encoded.
 799  *
 800  * Function switches to B encoding and encodeHeaderBase64() function, if
 801  * string is 8bit and multibyte character set supported by mbstring extension
 802  * is used. It can cause E_USER_NOTICE errors, if interface is used with
 803  * multibyte character set unsupported by mbstring extension.
 804  *
 805  * @param string $string header string, that has to be encoded
 806  * @return string quoted-printable encoded string
 807  * @todo make $mb_charsets system wide constant
 808  */
 809 function encodeHeader ($string) {
 810     global $default_charset, $languages, $squirrelmail_language;
 811
 812     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 813             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
 814         return  call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
 815     }
 816
 817     // Use B encoding for multibyte charsets
 818     $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
 819     if (in_array($default_charset,$mb_charsets) &&
 820         in_array($default_charset,sq_mb_list_encodings()) &&
 821         sq_is8bit($string)) {
 822         return encodeHeaderBase64($string,$default_charset);
 823     } elseif (in_array($default_charset,$mb_charsets) &&
 824               sq_is8bit($string) &&
 825               ! in_array($default_charset,sq_mb_list_encodings())) {
 826         // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
 827         // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
 828     }
 829
 830     // Encode only if the string contains 8-bit characters or =?
 831     $j = strlen($string);
 832     $max_l = 75 - strlen($default_charset) - 7;
 833     $aRet = array();
 834     $ret = '';
 835     $iEncStart = $enc_init = false;
 836     $cur_l = $iOffset = 0;
 837     for($i = 0; $i < $j; ++$i) {
 838         switch($string{$i})
 839         {
 840             case '=':
 841             case '<':
 842             case '>':
 843             case ',':
 844             case '?':
 845             case '_':
 846                 if ($iEncStart === false) {
 847                     $iEncStart = $i;
 848                 }
 849                 $cur_l+=3;
 850                 if ($cur_l > ($max_l-2)) {
 851                     /* if there is an stringpart that doesn't need encoding, add it */
 852                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 853                     $aRet[] = "=?$default_charset?Q?$ret?=";
 854                     $iOffset = $i;
 855                     $cur_l = 0;
 856                     $ret = '';
 857                     $iEncStart = false;
 858                 } else {
 859                     $ret .= sprintf("=%02X",ord($string{$i}));
 860                 }
 861                 break;
 862             case '(':
 863             case ')':
 864                 if ($iEncStart !== false) {
 865                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 866                     $aRet[] = "=?$default_charset?Q?$ret?=";
 867                     $iOffset = $i;
 868                     $cur_l = 0;
 869                     $ret = '';
 870                     $iEncStart = false;
 871                 }
 872                 break;
 873             case ' ':
 874                 if ($iEncStart !== false) {
 875                     $cur_l++;
 876                     if ($cur_l > $max_l) {
 877                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 878                         $aRet[] = "=?$default_charset?Q?$ret?=";
 879                         $iOffset = $i;
 880                         $cur_l = 0;
 881                         $ret = '';
 882                         $iEncStart = false;
 883                     } else {
 884                         $ret .= '_';
 885                     }
 886                 }
 887                 break;
 888             default:
 889                 $k = ord($string{$i});
 890                 if ($k > 126) {
 891                     if ($iEncStart === false) {
 892                         // do not start encoding in the middle of a string, also take the rest of the word.
 893                         $sLeadString = substr($string,0,$i);
 894                         $aLeadString = explode(' ',$sLeadString);
 895                         $sToBeEncoded = array_pop($aLeadString);
 896                         $iEncStart = $i - strlen($sToBeEncoded);
 897                         $ret .= $sToBeEncoded;
 898                         $cur_l += strlen($sToBeEncoded);
 899                     }
 900                     $cur_l += 3;
 901                     /* first we add the encoded string that reached it's max size */
 902                     if ($cur_l > ($max_l-2)) {
 903                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 904                         $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
 905                         $cur_l = 3;
 906                         $ret = '';
 907                         $iOffset = $i;
 908                         $iEncStart = $i;
 909                     }
 910                     $enc_init = true;
 911                     $ret .= sprintf("=%02X", $k);
 912                 } else {
 913                     if ($iEncStart !== false) {
 914                         $cur_l++;
 915                         if ($cur_l > $max_l) {
 916                             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 917                             $aRet[] = "=?$default_charset?Q?$ret?=";
 918                             $iEncStart = false;
 919                             $iOffset = $i;
 920                             $cur_l = 0;
 921                             $ret = '';
 922                         } else {
 923                             $ret .= $string{$i};
 924                         }
 925                     }
 926                 }
 927                 break;
 928         }
 929     }
 930
 931     if ($enc_init) {
 932         if ($iEncStart !== false) {
 933             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 934             $aRet[] = "=?$default_charset?Q?$ret?=";
 935         } else {
 936             $aRet[] = substr($string,$iOffset);
 937         }
 938         $string = implode('',$aRet);
 939     }
 940     return $string;
 941 }
 942
 943 /**
 944  * Encodes string according to rfc2047 B encoding header formating rules
 945  *
 946  * It is recommended way to encode headers with character sets that store
 947  * symbols in more than one byte.
 948  *
 949  * Function requires mbstring support. If required mbstring functions are missing,
 950  * function returns false and sets E_USER_WARNING level error message.
 951  *
 952  * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
 953  * that mbstring functions will generate E_WARNING errors, if unsupported
 954  * character set is used. mb_encode_mimeheader function provided by php
 955  * mbstring extension is not used in order to get better control of header
 956  * encoding.
 957  *
 958  * Used php code functions - function_exists(), trigger_error(), strlen()
 959  * (is used with charset names and base64 strings). Used php mbstring
 960  * functions - mb_strlen and mb_substr.
 961  *
 962  * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
 963  * encoding), rfc 2822 (header folding)
 964  *
 965  * @param string $string header string that must be encoded
 966  * @param string $charset character set. Must be supported by mbstring extension.
 967  * Use sq_mb_list_encodings() to detect supported charsets.
 968  * @return string string encoded according to rfc2047 B encoding formating rules
 969  * @since 1.5.1
 970  * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
 971  * @todo Do we want to control max length of header?
 972  * @todo Do we want to control EOL (end-of-line) marker?
 973  * @todo Do we want to translate error message?
 974  */
 975 function encodeHeaderBase64($string,$charset) {
 976     /**
 977      * Check mbstring function requirements.
 978      */
 979     if (! function_exists('mb_strlen') ||
 980         ! function_exists('mb_substr')) {
 981         // set E_USER_WARNING
 982         trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
 983         // return false
 984         return false;
 985     }
 986
 987     // initial return array
 988     $aRet = array();
 989
 990     /**
 991      * header length = 75 symbols max (same as in encodeHeader)
 992      * remove $charset length
 993      * remove =? ? ?= (5 chars)
 994      * remove 2 more chars (\r\n ?)
 995      */
 996     $iMaxLength = 75 - strlen($charset) - 7;
 997
 998     // set first character position
 999     $iStartCharNum = 0;
1000
1001     // loop through all characters. count characters and not bytes.
1002     for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1003         // encode string from starting character to current character.
1004         $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1005
1006         // Check encoded string length
1007         if(strlen($encoded_string)>$iMaxLength) {
1008             // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1009             $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1010
1011             // set new starting character
1012             $iStartCharNum = $iCharNum-1;
1013
1014             // encode last char (in case it is last character in string)
1015             $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1016         } // if string is shorter than max length - add next character
1017     }
1018
1019     // add last encoded string to array
1020     $aRet[] = $encoded_string;
1021
1022     // set initial return string
1023     $sRet = '';
1024
1025     // loop through encoded strings
1026     foreach($aRet as $string) {
1027         // TODO: Do we want to control EOL (end-of-line) marker
1028         if ($sRet!='') $sRet.= " ";
1029
1030         // add header tags and encoded string to return string
1031         $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1032     }
1033
1034     return $sRet;
1035 }
1036
1037 /* This function trys to locate the entity_id of a specific mime element */
1038 function find_ent_id($id, $message) {
1039     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1040         if ($message->entities[$i]->header->type0 == 'multipart')  {
1041             $ret = find_ent_id($id, $message->entities[$i]);
1042         } else {
1043             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
1044 //                if (sq_check_save_extension($message->entities[$i])) {
1045                 return $message->entities[$i]->entity_id;
1046 //                }
1047             } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1048                 /**
1049                  * This is part of a fix for Outlook Express 6.x generating
1050                  * cid URLs without creating content-id headers
1051                  * @@JA - 20050207
1052                  */
1053                 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1054                     return $message->entities[$i]->entity_id;
1055                 }
1056             }
1057         }
1058     }
1059     return $ret;
1060 }
1061
1062 function sq_check_save_extension($message) {
1063     $filename = $message->getFilename();
1064     $ext = substr($filename, strrpos($filename,'.')+1);
1065     $save_extensions = array('jpg','jpeg','gif','png','bmp');
1066     return in_array($ext, $save_extensions);
1067 }
1068
1069
1070 /**
1071  ** HTMLFILTER ROUTINES
1072  */
1073
1074 /**
1075  * This function checks attribute values for entity-encoded values
1076  * and returns them translated into 8-bit strings so we can run
1077  * checks on them.
1078  *
1079  * @param  $attvalue A string to run entity check against.
1080  * @return           Nothing, modifies a reference value.
1081  */
1082 function sq_defang(&$attvalue){
1083     $me = 'sq_defang';
1084     /**
1085      * Skip this if there aren't ampersands or backslashes.
1086      */
1087     if (strpos($attvalue, '&') === false
1088         && strpos($attvalue, '\\') === false){
1089         return;
1090     }
1091     $m = false;
1092     do {
1093         $m = false;
1094         $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1095         $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1096         $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1097     } while ($m == true);
1098     $attvalue = stripslashes($attvalue);
1099 }
1100
1101 /**
1102  * Kill any tabs, newlines, or carriage returns. Our friends the
1103  * makers of the browser with 95% market value decided that it'd
1104  * be funny to make "java[tab]script" be just as good as "javascript".
1105  *
1106  * @param  attvalue  The attribute value before extraneous spaces removed.
1107  * @return attvalue  Nothing, modifies a reference value.
1108  */
1109 function sq_unspace(&$attvalue){
1110     $me = 'sq_unspace';
1111     if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1112         $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1113                                 Array('',   '',   '',   '',   ''), $attvalue);
1114     }
1115 }
1116
1117 /**
1118  * This function returns the final tag out of the tag name, an array
1119  * of attributes, and the type of the tag. This function is called by
1120  * sq_sanitize internally.
1121  *
1122  * @param  $tagname  the name of the tag.
1123  * @param  $attary   the array of attributes and their values
1124  * @param  $tagtype  The type of the tag (see in comments).
1125  * @return           a string with the final tag representation.
1126  */
1127 function sq_tagprint($tagname, $attary, $tagtype){
1128     $me = 'sq_tagprint';
1129
1130     if ($tagtype == 2){
1131         $fulltag = '</' . $tagname . '>';
1132     } else {
1133         $fulltag = '<' . $tagname;
1134         if (is_array($attary) && sizeof($attary)){
1135             $atts = Array();
1136             while (list($attname, $attvalue) = each($attary)){
1137                 array_push($atts, "$attname=$attvalue");
1138             }
1139             $fulltag .= ' ' . join(" ", $atts);
1140         }
1141         if ($tagtype == 3){
1142             $fulltag .= ' /';
1143         }
1144         $fulltag .= '>';
1145     }
1146     return $fulltag;
1147 }
1148
1149 /**
1150  * A small helper function to use with array_walk. Modifies a by-ref
1151  * value and makes it lowercase.
1152  *
1153  * @param  $val a value passed by-ref.
1154  * @return      void since it modifies a by-ref value.
1155  */
1156 function sq_casenormalize(&$val){
1157     $val = strtolower($val);
1158 }
1159
1160 /**
1161  * This function skips any whitespace from the current position within
1162  * a string and to the next non-whitespace value.
1163  *
1164  * @param  $body   the string
1165  * @param  $offset the offset within the string where we should start
1166  *                 looking for the next non-whitespace character.
1167  * @return         the location within the $body where the next
1168  *                 non-whitespace char is located.
1169  */
1170 function sq_skipspace($body, $offset){
1171     $me = 'sq_skipspace';
1172     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1173     if (sizeof($matches{1})){
1174         $count = strlen($matches{1});
1175         $offset += $count;
1176     }
1177     return $offset;
1178 }
1179
1180 /**
1181  * This function looks for the next character within a string.  It's
1182  * really just a glorified "strpos", except it catches if failures
1183  * nicely.
1184  *
1185  * @param  $body   The string to look for needle in.
1186  * @param  $offset Start looking from this position.
1187  * @param  $needle The character/string to look for.
1188  * @return         location of the next occurance of the needle, or
1189  *                 strlen($body) if needle wasn't found.
1190  */
1191 function sq_findnxstr($body, $offset, $needle){
1192     $me  = 'sq_findnxstr';
1193     $pos = strpos($body, $needle, $offset);
1194     if ($pos === FALSE){
1195         $pos = strlen($body);
1196     }
1197     return $pos;
1198 }
1199
1200 /**
1201  * This function takes a PCRE-style regexp and tries to match it
1202  * within the string.
1203  *
1204  * @param  $body   The string to look for needle in.
1205  * @param  $offset Start looking from here.
1206  * @param  $reg    A PCRE-style regex to match.
1207  * @return         Returns a false if no matches found, or an array
1208  *                 with the following members:
1209  *                 - integer with the location of the match within $body
1210  *                 - string with whatever content between offset and the match
1211  *                 - string with whatever it is we matched
1212  */
1213 function sq_findnxreg($body, $offset, $reg){
1214     $me = 'sq_findnxreg';
1215     $matches = Array();
1216     $retarr = Array();
1217     preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1218     if (!isset($matches{0}) || !$matches{0}){
1219         $retarr = false;
1220     } else {
1221         $retarr{0} = $offset + strlen($matches{1});
1222         $retarr{1} = $matches{1};
1223         $retarr{2} = $matches{2};
1224     }
1225     return $retarr;
1226 }
1227
1228 /**
1229  * This function looks for the next tag.
1230  *
1231  * @param  $body   String where to look for the next tag.
1232  * @param  $offset Start looking from here.
1233  * @return         false if no more tags exist in the body, or
1234  *                 an array with the following members:
1235  *                 - string with the name of the tag
1236  *                 - array with attributes and their values
1237  *                 - integer with tag type (1, 2, or 3)
1238  *                 - integer where the tag starts (starting "<")
1239  *                 - integer where the tag ends (ending ">")
1240  *                 first three members will be false, if the tag is invalid.
1241  */
1242 function sq_getnxtag($body, $offset){
1243     $me = 'sq_getnxtag';
1244     if ($offset > strlen($body)){
1245         return false;
1246     }
1247     $lt = sq_findnxstr($body, $offset, "<");
1248     if ($lt == strlen($body)){
1249         return false;
1250     }
1251     /**
1252      * We are here:
1253      * blah blah <tag attribute="value">
1254      * \---------^
1255      */
1256     $pos = sq_skipspace($body, $lt+1);
1257     if ($pos >= strlen($body)){
1258         return Array(false, false, false, $lt, strlen($body));
1259     }
1260     /**
1261      * There are 3 kinds of tags:
1262      * 1. Opening tag, e.g.:
1263      *    <a href="blah">
1264      * 2. Closing tag, e.g.:
1265      *    </a>
1266      * 3. XHTML-style content-less tag, e.g.:
1267      *    <img src="blah" />
1268      */
1269     $tagtype = false;
1270     switch (substr($body, $pos, 1)){
1271         case '/':
1272             $tagtype = 2;
1273             $pos++;
1274             break;
1275         case '!':
1276             /**
1277              * A comment or an SGML declaration.
1278              */
1279             if (substr($body, $pos+1, 2) == "--"){
1280                 $gt = strpos($body, "-->", $pos);
1281                 if ($gt === false){
1282                     $gt = strlen($body);
1283                 } else {
1284                     $gt += 2;
1285                 }
1286                 return Array(false, false, false, $lt, $gt);
1287             } else {
1288                 $gt = sq_findnxstr($body, $pos, ">");
1289                 return Array(false, false, false, $lt, $gt);
1290             }
1291             break;
1292         default:
1293             /**
1294              * Assume tagtype 1 for now. If it's type 3, we'll switch values
1295              * later.
1296              */
1297             $tagtype = 1;
1298             break;
1299     }
1300
1301     $tag_start = $pos;
1302     $tagname = '';
1303     /**
1304      * Look for next [\W-_], which will indicate the end of the tag name.
1305      */
1306     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1307     if ($regary == false){
1308         return Array(false, false, false, $lt, strlen($body));
1309     }
1310     list($pos, $tagname, $match) = $regary;
1311     $tagname = strtolower($tagname);
1312
1313     /**
1314      * $match can be either of these:
1315      * '>'  indicating the end of the tag entirely.
1316      * '\s' indicating the end of the tag name.
1317      * '/'  indicating that this is type-3 xhtml tag.
1318      *
1319      * Whatever else we find there indicates an invalid tag.
1320      */
1321     switch ($match){
1322         case '/':
1323             /**
1324              * This is an xhtml-style tag with a closing / at the
1325              * end, like so: <img src="blah" />. Check if it's followed
1326              * by the closing bracket. If not, then this tag is invalid
1327              */
1328             if (substr($body, $pos, 2) == "/>"){
1329                 $pos++;
1330                 $tagtype = 3;
1331             } else {
1332                 $gt = sq_findnxstr($body, $pos, ">");
1333                 $retary = Array(false, false, false, $lt, $gt);
1334                 return $retary;
1335             }
1336         case '>':
1337             return Array($tagname, false, $tagtype, $lt, $pos);
1338             break;
1339         default:
1340             /**
1341              * Check if it's whitespace
1342              */
1343             if (!preg_match('/\s/', $match)){
1344                 /**
1345                  * This is an invalid tag! Look for the next closing ">".
1346                  */
1347                 $gt = sq_findnxstr($body, $lt, ">");
1348                 return Array(false, false, false, $lt, $gt);
1349             }
1350             break;
1351     }
1352
1353     /**
1354      * At this point we're here:
1355      * <tagname  attribute='blah'>
1356      * \-------^
1357      *
1358      * At this point we loop in order to find all attributes.
1359      */
1360     $attname = '';
1361     $atttype = false;
1362     $attary = Array();
1363
1364     while ($pos <= strlen($body)){
1365         $pos = sq_skipspace($body, $pos);
1366         if ($pos == strlen($body)){
1367             /**
1368              * Non-closed tag.
1369              */
1370             return Array(false, false, false, $lt, $pos);
1371         }
1372         /**
1373          * See if we arrived at a ">" or "/>", which means that we reached
1374          * the end of the tag.
1375          */
1376         $matches = Array();
1377         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1378             /**
1379              * Yep. So we did.
1380              */
1381             $pos += strlen($matches{1});
1382             if ($matches{2} == "/>"){
1383                 $tagtype = 3;
1384                 $pos++;
1385             }
1386             return Array($tagname, $attary, $tagtype, $lt, $pos);
1387         }
1388
1389         /**
1390          * There are several types of attributes, with optional
1391          * [:space:] between members.
1392          * Type 1:
1393          *   attrname[:space:]=[:space:]'CDATA'
1394          * Type 2:
1395          *   attrname[:space:]=[:space:]"CDATA"
1396          * Type 3:
1397          *   attr[:space:]=[:space:]CDATA
1398          * Type 4:
1399          *   attrname
1400          *
1401          * We leave types 1 and 2 the same, type 3 we check for
1402          * '"' and convert to "&quot" if needed, then wrap in
1403          * double quotes. Type 4 we convert into:
1404          * attrname="yes".
1405          */
1406         $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1407         if ($regary == false){
1408             /**
1409              * Looks like body ended before the end of tag.
1410              */
1411             return Array(false, false, false, $lt, strlen($body));
1412         }
1413         list($pos, $attname, $match) = $regary;
1414         $attname = strtolower($attname);
1415         /**
1416          * We arrived at the end of attribute name. Several things possible
1417          * here:
1418          * '>'  means the end of the tag and this is attribute type 4
1419          * '/'  if followed by '>' means the same thing as above
1420          * '\s' means a lot of things -- look what it's followed by.
1421          *      anything else means the attribute is invalid.
1422          */
1423         switch($match){
1424             case '/':
1425                 /**
1426                  * This is an xhtml-style tag with a closing / at the
1427                  * end, like so: <img src="blah" />. Check if it's followed
1428                  * by the closing bracket. If not, then this tag is invalid
1429                  */
1430                 if (substr($body, $pos, 2) == "/>"){
1431                     $pos++;
1432                     $tagtype = 3;
1433                 } else {
1434                     $gt = sq_findnxstr($body, $pos, ">");
1435                     $retary = Array(false, false, false, $lt, $gt);
1436                     return $retary;
1437                 }
1438             case '>':
1439                 $attary{$attname} = '"yes"';
1440                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1441                 break;
1442             default:
1443                 /**
1444                  * Skip whitespace and see what we arrive at.
1445                  */
1446                 $pos = sq_skipspace($body, $pos);
1447                 $char = substr($body, $pos, 1);
1448                 /**
1449                  * Two things are valid here:
1450                  * '=' means this is attribute type 1 2 or 3.
1451                  * \w means this was attribute type 4.
1452                  * anything else we ignore and re-loop. End of tag and
1453                  * invalid stuff will be caught by our checks at the beginning
1454                  * of the loop.
1455                  */
1456                 if ($char == "="){
1457                     $pos++;
1458                     $pos = sq_skipspace($body, $pos);
1459                     /**
1460                      * Here are 3 possibilities:
1461                      * "'"  attribute type 1
1462                      * '"'  attribute type 2
1463                      * everything else is the content of tag type 3
1464                      */
1465                     $quot = substr($body, $pos, 1);
1466                     if ($quot == "'"){
1467                         $regary = sq_findnxreg($body, $pos+1, "\'");
1468                         if ($regary == false){
1469                             return Array(false, false, false, $lt, strlen($body));
1470                         }
1471                         list($pos, $attval, $match) = $regary;
1472                         $pos++;
1473                         $attary{$attname} = "'" . $attval . "'";
1474                     } else if ($quot == '"'){
1475                         $regary = sq_findnxreg($body, $pos+1, '\"');
1476                         if ($regary == false){
1477                             return Array(false, false, false, $lt, strlen($body));
1478                         }
1479                         list($pos, $attval, $match) = $regary;
1480                         $pos++;
1481                         $attary{$attname} = '"' . $attval . '"';
1482                     } else {
1483                         /**
1484                          * These are hateful. Look for \s, or >.
1485                          */
1486                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1487                         if ($regary == false){
1488                             return Array(false, false, false, $lt, strlen($body));
1489                         }
1490                         list($pos, $attval, $match) = $regary;
1491                         /**
1492                          * If it's ">" it will be caught at the top.
1493                          */
1494                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1495                         $attary{$attname} = '"' . $attval . '"';
1496                     }
1497                 } else if (preg_match("|[\w/>]|", $char)) {
1498                     /**
1499                      * That was attribute type 4.
1500                      */
1501                     $attary{$attname} = '"yes"';
1502                 } else {
1503                     /**
1504                      * An illegal character. Find next '>' and return.
1505                      */
1506                     $gt = sq_findnxstr($body, $pos, ">");
1507                     return Array(false, false, false, $lt, $gt);
1508                 }
1509                 break;
1510         }
1511     }
1512     /**
1513      * The fact that we got here indicates that the tag end was never
1514      * found. Return invalid tag indication so it gets stripped.
1515      */
1516     return Array(false, false, false, $lt, strlen($body));
1517 }
1518
1519 /**
1520  * Translates entities into literal values so they can be checked.
1521  *
1522  * @param $attvalue the by-ref value to check.
1523  * @param $regex    the regular expression to check against.
1524  * @param $hex      whether the entites are hexadecimal.
1525  * @return          True or False depending on whether there were matches.
1526  */
1527 function sq_deent(&$attvalue, $regex, $hex=false){
1528     $me = 'sq_deent';
1529     $ret_match = false;
1530     preg_match_all($regex, $attvalue, $matches);
1531     if (is_array($matches) && sizeof($matches[0]) > 0){
1532         $repl = Array();
1533         for ($i = 0; $i < sizeof($matches[0]); $i++){
1534             $numval = $matches[1][$i];
1535             if ($hex){
1536                 $numval = hexdec($numval);
1537             }
1538             $repl{$matches[0][$i]} = chr($numval);
1539         }
1540         $attvalue = strtr($attvalue, $repl);
1541         return true;
1542     } else {
1543         return false;
1544     }
1545 }
1546
1547 /**
1548  * This function runs various checks against the attributes.
1549  *
1550  * @param  $tagname         String with the name of the tag.
1551  * @param  $attary          Array with all tag attributes.
1552  * @param  $rm_attnames     See description for sq_sanitize
1553  * @param  $bad_attvals     See description for sq_sanitize
1554  * @param  $add_attr_to_tag See description for sq_sanitize
1555  * @param  $message         message object
1556  * @param  $id              message id
1557  * @return                  Array with modified attributes.
1558  */
1559 function sq_fixatts($tagname,
1560                     $attary,
1561                     $rm_attnames,
1562                     $bad_attvals,
1563                     $add_attr_to_tag,
1564                     $message,
1565                     $id,
1566                     $mailbox
1567                     ){
1568     $me = 'sq_fixatts';
1569     while (list($attname, $attvalue) = each($attary)){
1570         /**
1571          * See if this attribute should be removed.
1572          */
1573         foreach ($rm_attnames as $matchtag=>$matchattrs){
1574             if (preg_match($matchtag, $tagname)){
1575                 foreach ($matchattrs as $matchattr){
1576                     if (preg_match($matchattr, $attname)){
1577                         unset($attary{$attname});
1578                         continue;
1579                     }
1580                 }
1581             }
1582         }
1583         /**
1584          * Remove any backslashes, entities, and extraneous whitespace.
1585          */
1586         sq_defang($attvalue);
1587         sq_unspace($attvalue);
1588
1589         /**
1590          * Now let's run checks on the attvalues.
1591          * I don't expect anyone to comprehend this. If you do,
1592          * get in touch with me so I can drive to where you live and
1593          * shake your hand personally. :)
1594          */
1595         foreach ($bad_attvals as $matchtag=>$matchattrs){
1596             if (preg_match($matchtag, $tagname)){
1597                 foreach ($matchattrs as $matchattr=>$valary){
1598                     if (preg_match($matchattr, $attname)){
1599                         /**
1600                          * There are two arrays in valary.
1601                          * First is matches.
1602                          * Second one is replacements
1603                          */
1604                         list($valmatch, $valrepl) = $valary;
1605                         $newvalue =
1606                             preg_replace($valmatch, $valrepl, $attvalue);
1607                         if ($newvalue != $attvalue){
1608                             $attary{$attname} = $newvalue;
1609                         }
1610                     }
1611                 }
1612             }
1613         }
1614
1615         /**
1616          * Replace empty src tags with the blank image.  src is only used
1617          * for frames, images, and image inputs.  Doing a replace should
1618          * not affect them working as should be, however it will stop
1619          * IE from being kicked off when src for img tags are not set
1620          */
1621         if (($attname == 'src') && ($attvalue == '""')) {
1622             $attary{$attname} = '"' . SM_PATH . 'images/blank.png"';
1623         }
1624
1625         /**
1626          * Turn cid: urls into http-friendly ones.
1627          */
1628         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1629             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1630         }
1631
1632         /**
1633          * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1634          * One day MS might actually make it match something useful, for now, falling
1635          * back to using cid2http, so we can grab the blank.png.
1636          */
1637         if (preg_match("/^[\'\"]\s*outbind:\/\//si", $attvalue)) {
1638             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1639         }
1640
1641     }
1642     /**
1643      * See if we need to append any attributes to this tag.
1644      */
1645     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1646         if (preg_match($matchtag, $tagname)){
1647             $attary = array_merge($attary, $addattary);
1648         }
1649     }
1650     return $attary;
1651 }
1652
1653 /**
1654  * This function edits the style definition to make them friendly and
1655  * usable in SquirrelMail.
1656  *
1657  * @param  $message  the message object
1658  * @param  $id       the message id
1659  * @param  $content  a string with whatever is between <style> and </style>
1660  * @param  $mailbox  the message mailbox
1661  * @return           a string with edited content.
1662  */
1663 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1664     global $view_unsafe_images;
1665     $me = 'sq_fixstyle';
1666     $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1667     if ($ret == FALSE){
1668         return array(FALSE, strlen($body));
1669     }
1670     $newpos = $ret[0] + strlen($ret[2]);
1671     $content = $ret[1];
1672     /**
1673     * First look for general BODY style declaration, which would be
1674     * like so:
1675     * body {background: blah-blah}
1676     * and change it to .bodyclass so we can just assign it to a <div>
1677     */
1678     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1679     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1680     /**
1681     * Fix url('blah') declarations.
1682     */
1683     //   $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1684     //                           "url(\\1$secremoveimg\\2)", $content);
1685     // remove NUL
1686     $content = str_replace("\0", "", $content);
1687     // NB I insert NUL characters to keep to avoid an infinite loop. They are removed after the loop.
1688     while (preg_match("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si", $content, $matches)) {
1689         $sProto = strtolower($matches[1]);
1690         switch ($sProto) {
1691           /**
1692            * Fix url('https*://.*) declarations but only if $view_unsafe_images
1693            * is false.
1694            */
1695           case 'https':
1696           case 'http':
1697             if (!$view_unsafe_images){
1698                 $sExpr = "/url\s*\(\s*([\'\"])\s*$sProto*:.*?([\'\"])\s*\)/si";
1699                 $content = preg_replace($sExpr, "u\0r\0l(\\1$secremoveimg\\2)", $content);
1700             }
1701             break;
1702           /**
1703            * Fix urls that refer to cid:
1704            */
1705           case 'cid':
1706             $cidurl = 'cid:'. $matches[2];
1707             $httpurl = sq_cid2http($message, $id, $cidurl, $mailbox);
1708             $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1709                                 "u\0r\0l($httpurl)", $content);
1710             break;
1711           default:
1712             /**
1713              * replace url with protocol other then the white list
1714              * http,https and cid by an empty string.
1715              */
1716             $content = preg_replace("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si",
1717                                 "", $content);
1718             break;
1719         }
1720         break;
1721     }
1722     // remove NUL
1723     $content = str_replace("\0", "", $content);
1724
1725    /**
1726     * Remove any backslashes, entities, and extraneous whitespace.
1727     */
1728     $contentTemp = $content;
1729     sq_defang($contentTemp);
1730     sq_unspace($contentTemp);
1731
1732     /**
1733      * Fix stupid css declarations which lead to vulnerabilities
1734      * in IE.
1735      */
1736     $match   = Array('/expression/i',
1737                     '/behaviou*r/i',
1738                     '/binding/i',
1739                     '/include-source/i');
1740     $replace = Array('idiocy', 'idiocy', 'idiocy', 'idiocy');
1741     $contentNew = preg_replace($match, $replace, $contentTemp);
1742     if ($contentNew !== $contentTemp) {
1743         // insecure css declarations are used. From now on we don't care
1744         // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
1745         $content = $contentNew;
1746     }
1747     return array($content, $newpos);
1748 }
1749
1750
1751 /**
1752  * This function converts cid: url's into the ones that can be viewed in
1753  * the browser.
1754  *
1755  * @param  $message  the message object
1756  * @param  $id       the message id
1757  * @param  $cidurl   the cid: url.
1758  * @param  $mailbox  the message mailbox
1759  * @return           a string with a http-friendly url
1760  */
1761 function sq_cid2http($message, $id, $cidurl, $mailbox){
1762     /**
1763      * Get rid of quotes.
1764      */
1765     $quotchar = substr($cidurl, 0, 1);
1766     if ($quotchar == '"' || $quotchar == "'"){
1767         $cidurl = str_replace($quotchar, "", $cidurl);
1768     } else {
1769         $quotchar = '';
1770     }
1771     $cidurl = substr(trim($cidurl), 4);
1772
1773     $match_str = '/\{.*?\}\//';
1774     $str_rep = '';
1775     $cidurl = preg_replace($match_str, $str_rep, $cidurl);
1776
1777     $linkurl = find_ent_id($cidurl, $message);
1778     /* in case of non-save cid links $httpurl should be replaced by a sort of
1779        unsave link image */
1780     $httpurl = '';
1781
1782     /**
1783      * This is part of a fix for Outlook Express 6.x generating
1784      * cid URLs without creating content-id headers. These images are
1785      * not part of the multipart/related html mail. The html contains
1786      * <img src="cid:{some_id}/image_filename.ext"> references to
1787      * attached images with as goal to render them inline although
1788      * the attachment disposition property is not inline.
1789      */
1790
1791     if (empty($linkurl)) {
1792         if (preg_match('/{.*}\//', $cidurl)) {
1793             $cidurl = preg_replace('/{.*}\//','', $cidurl);
1794             if (!empty($cidurl)) {
1795                 $linkurl = find_ent_id($cidurl, $message);
1796             }
1797         }
1798     }
1799
1800     if (!empty($linkurl)) {
1801         $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1802             "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1803             '&amp;ent_id=' . $linkurl . $quotchar;
1804     } else {
1805         /**
1806          * If we couldn't generate a proper img url, drop in a blank image
1807          * instead of sending back empty, otherwise it causes unusual behaviour
1808          */
1809         $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
1810     }
1811
1812     return $httpurl;
1813 }
1814
1815 /**
1816  * This function changes the <body> tag into a <div> tag since we
1817  * can't really have a body-within-body.
1818  *
1819  * @param  $attary   an array of attributes and values of <body>
1820  * @param  $mailbox  mailbox we're currently reading (for cid2http)
1821  * @param  $message  current message (for cid2http)
1822  * @param  $id       current message id (for cid2http)
1823  * @return           a modified array of attributes to be set for <div>
1824  */
1825 function sq_body2div($attary, $mailbox, $message, $id){
1826     $me = 'sq_body2div';
1827     $divattary = Array('class' => "'bodyclass'");
1828     $text = '#000000';
1829     $has_bgc_stl = $has_txt_stl = false;
1830     $styledef = '';
1831     if (is_array($attary) && sizeof($attary) > 0){
1832         foreach ($attary as $attname=>$attvalue){
1833             $quotchar = substr($attvalue, 0, 1);
1834             $attvalue = str_replace($quotchar, "", $attvalue);
1835             switch ($attname){
1836                 case 'background':
1837                     $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1838                     $styledef .= "background-image: url('$attvalue'); ";
1839                     break;
1840                 case 'bgcolor':
1841                     $has_bgc_stl = true;
1842                     $styledef .= "background-color: $attvalue; ";
1843                     break;
1844                 case 'text':
1845                     $has_txt_stl = true;
1846                     $styledef .= "color: $attvalue; ";
1847                     break;
1848             }
1849         }
1850         // Outlook defines a white bgcolor and no text color. This can lead to
1851         // white text on a white bg with certain themes.
1852         if ($has_bgc_stl && !$has_txt_stl) {
1853             $styledef .= "color: $text; ";
1854         }
1855         if (strlen($styledef) > 0){
1856             $divattary{"style"} = "\"$styledef\"";
1857         }
1858     }
1859     return $divattary;
1860 }
1861
1862 /**
1863  * This is the main function and the one you should actually be calling.
1864  * There are several variables you should be aware of an which need
1865  * special description.
1866  *
1867  * Since the description is quite lengthy, see it here:
1868  * http://linux.duke.edu/projects/mini/htmlfilter/
1869  *
1870  * @param $body                 the string with HTML you wish to filter
1871  * @param $tag_list             see description above
1872  * @param $rm_tags_with_content see description above
1873  * @param $self_closing_tags    see description above
1874  * @param $force_tag_closing    see description above
1875  * @param $rm_attnames          see description above
1876  * @param $bad_attvals          see description above
1877  * @param $add_attr_to_tag      see description above
1878  * @param $message              message object
1879  * @param $id                   message id
1880  * @return                      sanitized html safe to show on your pages.
1881  */
1882 function sq_sanitize($body,
1883                      $tag_list,
1884                      $rm_tags_with_content,
1885                      $self_closing_tags,
1886                      $force_tag_closing,
1887                      $rm_attnames,
1888                      $bad_attvals,
1889                      $add_attr_to_tag,
1890                      $message,
1891                      $id,
1892                      $mailbox
1893                      ){
1894     $me = 'sq_sanitize';
1895     $rm_tags = array_shift($tag_list);
1896     /**
1897      * Normalize rm_tags and rm_tags_with_content.
1898      */
1899     @array_walk($tag_list, 'sq_casenormalize');
1900     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1901     @array_walk($self_closing_tags, 'sq_casenormalize');
1902     /**
1903      * See if tag_list is of tags to remove or tags to allow.
1904      * false  means remove these tags
1905      * true   means allow these tags
1906      */
1907     $curpos = 0;
1908     $open_tags = Array();
1909     $trusted = "\n<!-- begin sanitized html -->\n";
1910     $skip_content = false;
1911     /**
1912      * Take care of netscape's stupid javascript entities like
1913      * &{alert('boo')};
1914      */
1915     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1916
1917     while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1918         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1919         $free_content = substr($body, $curpos, $lt-$curpos);
1920         /**
1921          * Take care of <style>
1922          */
1923         if ($tagname == "style" && $tagtype == 1){
1924             list($free_content, $curpos) =
1925                 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
1926             if ($free_content != FALSE){
1927                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1928                 $trusted .= $free_content;
1929                 $trusted .= sq_tagprint($tagname, false, 2);
1930             }
1931             continue;
1932         }
1933         if ($skip_content == false){
1934             $trusted .= $free_content;
1935         }
1936         if ($tagname != FALSE){
1937             if ($tagtype == 2){
1938                 if ($skip_content == $tagname){
1939                     /**
1940                      * Got to the end of tag we needed to remove.
1941                      */
1942                     $tagname = false;
1943                     $skip_content = false;
1944                 } else {
1945                     if ($skip_content == false){
1946                         if ($tagname == "body"){
1947                             $tagname = "div";
1948                         }
1949                         if (isset($open_tags{$tagname}) &&
1950                                 $open_tags{$tagname} > 0){
1951                             $open_tags{$tagname}--;
1952                         } else {
1953                             $tagname = false;
1954                         }
1955                     }
1956                 }
1957             } else {
1958                 /**
1959                  * $rm_tags_with_content
1960                  */
1961                 if ($skip_content == false){
1962                     /**
1963                      * See if this is a self-closing type and change
1964                      * tagtype appropriately.
1965                      */
1966                     if ($tagtype == 1
1967                             && in_array($tagname, $self_closing_tags)){
1968                         $tagtype = 3;
1969                     }
1970                     /**
1971                      * See if we should skip this tag and any content
1972                      * inside it.
1973                      */
1974                     if ($tagtype == 1 &&
1975                             in_array($tagname, $rm_tags_with_content)){
1976                         $skip_content = $tagname;
1977                     } else {
1978                         if (($rm_tags == false
1979                                     && in_array($tagname, $tag_list)) ||
1980                                 ($rm_tags == true &&
1981                                  !in_array($tagname, $tag_list))){
1982                             $tagname = false;
1983                         } else {
1984                             /**
1985                              * Convert body into div.
1986                              */
1987                             if ($tagname == "body"){
1988                                 $tagname = "div";
1989                                 $attary = sq_body2div($attary, $mailbox,
1990                                         $message, $id);
1991                             }
1992                             if ($tagtype == 1){
1993                                 if (isset($open_tags{$tagname})){
1994                                     $open_tags{$tagname}++;
1995                                 } else {
1996                                     $open_tags{$tagname}=1;
1997                                 }
1998                             }
1999                             /**
2000                              * This is where we run other checks.
2001                              */
2002                             if (is_array($attary) && sizeof($attary) > 0){
2003                                 $attary = sq_fixatts($tagname,
2004                                                      $attary,
2005                                                      $rm_attnames,
2006                                                      $bad_attvals,
2007                                                      $add_attr_to_tag,
2008                                                      $message,
2009                                                      $id,
2010                                                      $mailbox
2011                                                      );
2012                             }
2013                         }
2014                     }
2015                 }
2016             }
2017             if ($tagname != false && $skip_content == false){
2018                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2019             }
2020         }
2021         $curpos = $gt+1;
2022     }
2023     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2024     if ($force_tag_closing == true){
2025         foreach ($open_tags as $tagname=>$opentimes){
2026             while ($opentimes > 0){
2027                 $trusted .= '</' . $tagname . '>';
2028                 $opentimes--;
2029             }
2030         }
2031         $trusted .= "\n";
2032     }
2033     $trusted .= "<!-- end sanitized html -->\n";
2034     return $trusted;
2035 }
2036
2037 /**
2038  * This is a wrapper function to call html sanitizing routines.
2039  *
2040  * @param  $body  the body of the message
2041  * @param  $id    the id of the message
2042
2043  * @param  $message
2044  * @param  $mailbox
2045  * @param  boolean $take_mailto_links When TRUE, converts mailto: links
2046  *                                    into internal SM compose links
2047  *                                    (optional; default = TRUE)
2048  * @return        a string with html safe to display in the browser.
2049  */
2050 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2051
2052     require_once(SM_PATH . 'functions/url_parser.php');  // for $MailTo_PReg_Match
2053
2054     global $attachment_common_show_images, $view_unsafe_images,
2055            $has_unsafe_images;
2056     /**
2057      * Don't display attached images in HTML mode.
2058      */
2059     $attachment_common_show_images = false;
2060     $tag_list = Array(
2061             false,
2062             "object",
2063             "meta",
2064             "html",
2065             "head",
2066             "base",
2067             "link",
2068             "frame",
2069             "iframe",
2070             "plaintext",
2071             "marquee"
2072             );
2073
2074     $rm_tags_with_content = Array(
2075             "script",
2076             "applet",
2077             "embed",
2078             "title",
2079             "frameset",
2080             "xmp",
2081             "xml"
2082             );
2083
2084     $self_closing_tags =  Array(
2085             "img",
2086             "br",
2087             "hr",
2088             "input",
2089             "outbind"
2090             );
2091
2092     $force_tag_closing = true;
2093
2094     $rm_attnames = Array(
2095             "/.*/" =>
2096             Array(
2097                 "/target/i",
2098                 "/^on.*/i",
2099                 "/^dynsrc/i",
2100                 "/^data.*/i",
2101                 "/^lowsrc.*/i"
2102                 )
2103             );
2104
2105     $secremoveimg = "../images/" . _("sec_remove_eng.png");
2106     $bad_attvals = Array(
2107             "/.*/" =>
2108             Array(
2109                 "/^src|background/i" =>
2110                 Array(
2111                     Array(
2112                         "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2113                         "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2114                         "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2115                         ),
2116                     Array(
2117                         "\\1$secremoveimg\\2",
2118                         "\\1$secremoveimg\\2",
2119                         "\\1$secremoveimg\\2",
2120                         )
2121                     ),
2122                 "/^href|action/i" =>
2123                 Array(
2124                     Array(
2125                         "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2126                         "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2127                         "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2128                         ),
2129                     Array(
2130                         "\\1#\\1",
2131                         "\\1#\\1",
2132                         "\\1#\\1"
2133                         )
2134                     ),
2135         "/^style/i" =>
2136             Array(
2137                 Array(
2138                     "/expression/i",
2139                     "/binding/i",
2140                     "/behaviou*r/i",
2141                     "/include-source/i",
2142                     "/position\s*:\s*absolute/i",
2143                     "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2144                     "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2145                     "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2146                     "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2147                     ),
2148                 Array(
2149                     "idiocy",
2150                     "idiocy",
2151                     "idiocy",
2152                     "idiocy",
2153                     "",
2154                     "url(\\1#\\1)",
2155                     "url(\\1#\\1)",
2156                     "url(\\1#\\1)",
2157                     "\\1:url(\\2#\\3)"
2158                     )
2159                 )
2160             )
2161         );
2162     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
2163         $view_unsafe_images = false;
2164     }
2165     if (!$view_unsafe_images){
2166         /**
2167          * Remove any references to http/https if view_unsafe_images set
2168          * to false.
2169          */
2170         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2171                 '/^([\'\"])\s*https*:.*([\'\"])/si');
2172         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2173                 "\\1$secremoveimg\\1");
2174         array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2175                 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2176         array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2177                 "url(\\1$secremoveimg\\1)");
2178     }
2179
2180     $add_attr_to_tag = Array(
2181             "/^a$/i" =>
2182             Array('target'=>'"_blank"',
2183                 'title'=>'"'._("This external link will open in a new window").'"'
2184                 )
2185             );
2186     $trusted = sq_sanitize($body,
2187                            $tag_list,
2188                            $rm_tags_with_content,
2189                            $self_closing_tags,
2190                            $force_tag_closing,
2191                            $rm_attnames,
2192                            $bad_attvals,
2193                            $add_attr_to_tag,
2194                            $message,
2195                            $id,
2196                            $mailbox
2197                            );
2198     if (preg_match("|$secremoveimg|i", $trusted)){
2199         $has_unsafe_images = true;
2200     }
2201
2202     // we want to parse mailto's in HTML output, change to SM compose links
2203     // this is a modified version of code from url_parser.php... but Marc is
2204     // right: we need a better filtering implementation; adding this randomly
2205     // here is not a great solution
2206     //
2207     if ($take_mailto_links) {
2208         // parseUrl($trusted);   // this even parses URLs inside of tags... too aggressive
2209         global $MailTo_PReg_Match;
2210         $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1);
2211         if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2212             foreach ($regs[0] as $i => $mailto_before) {
2213                 $mailto_params = $regs[10][$i];
2214                 // get rid of any tailing quote since we have to add send_to to the end
2215                 //
2216                 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2217                     $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2218                 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2219                     $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2220
2221                 if ($regs[1][$i]) {    //if there is an email addr before '?', we need to merge it with the params
2222                     $to = 'to=' . $regs[1][$i];
2223                     if (strpos($mailto_params, 'to=') > -1)    //already a 'to='
2224                         $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2225                     else {
2226                         if ($mailto_params)    //already some params, append to them
2227                             $mailto_params .= '&amp;' . $to;
2228                         else
2229                             $mailto_params .= '?' . $to;
2230                     }
2231                 }
2232
2233                 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2234
2235                 // we'll already have target=_blank, no need to allow comp_in_new
2236                 // here (which would be a lot more work anyway)
2237                 //
2238                 global $compose_new_win;
2239                 $temp_comp_in_new = $compose_new_win;
2240                 $compose_new_win = 0;
2241                 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2242                 $compose_new_win = $temp_comp_in_new;
2243
2244                 // remove <a href=" and anything after the next quote (we only
2245                 // need the uri, not the link HTML) in compose uri
2246                 //
2247                 $comp_uri = substr($comp_uri, 9);
2248                 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2249                 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2250             }
2251         }
2252     }
2253
2254     return $trusted;
2255 }
2256
2257 /**
2258  * function SendDownloadHeaders - send file to the browser
2259  *
2260  * Original Source: SM core src/download.php
2261  * moved here to make it available to other code, and separate
2262  * front end from back end functionality.
2263  *
2264  * @param string $type0 first half of mime type
2265  * @param string $type1 second half of mime type
2266  * @param string $filename filename to tell the browser for downloaded file
2267  * @param boolean $force whether to force the download dialog to pop
2268  * @param optional integer $filesize send the Content-Header and length to the browser
2269  * @return void
2270  */
2271 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2272     global $languages, $squirrelmail_language;
2273     $isIE = $isIE6 = 0;
2274
2275     sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2276
2277     if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2278             strstr($HTTP_USER_AGENT, 'Opera') === false) {
2279         $isIE = 1;
2280     }
2281
2282     if (strstr($HTTP_USER_AGENT, 'compatible; MSIE 6') !== false &&
2283             strstr($HTTP_USER_AGENT, 'Opera') === false) {
2284         $isIE6 = 1;
2285     }
2286
2287     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2288             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2289         $filename =
2290             call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2291     } else {
2292         $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
2293     }
2294
2295     // A Pox on Microsoft and it's Internet Explorer!
2296     //
2297     // IE has lots of bugs with file downloads.
2298     // It also has problems with SSL.  Both of these cause problems
2299     // for us in this function.
2300     //
2301     // See this article on Cache Control headers and SSL
2302     // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2303     //
2304     // The best thing you can do for IE is to upgrade to the latest
2305     // version
2306     //set all the Cache Control Headers for IE
2307     if ($isIE) {
2308         $filename=rawurlencode($filename);
2309         header ("Pragma: public");
2310         header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2311         header ("Cache-Control: post-check=0, pre-check=0", false);
2312         header ("Cache-Control: private");
2313
2314         //set the inline header for IE, we'll add the attachment header later if we need it
2315         header ("Content-Disposition: inline; filename=$filename");
2316     }
2317
2318     if (!$force) {
2319         // Try to show in browser window
2320         header ("Content-Disposition: inline; filename=\"$filename\"");
2321         header ("Content-Type: $type0/$type1; name=\"$filename\"");
2322     } else {
2323         // Try to pop up the "save as" box
2324
2325         // IE makes this hard.  It pops up 2 save boxes, or none.
2326         // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2327         // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2328         // But, according to Microsoft, it is "RFC compliant but doesn't
2329         // take into account some deviations that allowed within the
2330         // specification."  Doesn't that mean RFC non-compliant?
2331         // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2332
2333         // all browsers need the application/octet-stream header for this
2334         header ("Content-Type: application/octet-stream; name=\"$filename\"");
2335
2336         // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2337         // Do not have quotes around filename, but that applied to
2338         // "attachment"... does it apply to inline too?
2339         header ("Content-Disposition: attachment; filename=\"$filename\"");
2340
2341         if ($isIE && !$isIE6) {
2342             // This combination seems to work mostly.  IE 5.5 SP 1 has
2343             // known issues (see the Microsoft Knowledge Base)
2344
2345             // This works for most types, but doesn't work with Word files
2346             header ("Content-Type: application/download; name=\"$filename\"");
2347
2348             // These are spares, just in case.  :-)
2349             //header("Content-Type: $type0/$type1; name=\"$filename\"");
2350             //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2351             //header("Content-Type: application/octet-stream; name=\"$filename\"");
2352         } else {
2353             // another application/octet-stream forces download for Netscape
2354             header ("Content-Type: application/octet-stream; name=\"$filename\"");
2355         }
2356     }
2357
2358     //send the content-length header if the calling function provides it
2359     if ($filesize > 0) {
2360         header("Content-Length: $filesize");
2361     }
2362
2363 }  // end fn SendDownloadHeaders
2364
2365 ?>