functions/mime.php

   1 <?php
   2
   3 /**
   4  * mime.php
   5  *
   6  * This contains the functions necessary to detect and decode MIME
   7  * messages.
   8  *
   9  * @copyright &copy; 1999-2005 The SquirrelMail Project Team
  10  * @license http://opensource.org/licenses/gpl-license.php GNU Public License
  11  * @version $Id$
  12  * @package squirrelmail
  13  */
  14
  15 /** The typical includes... */
  16 require_once(SM_PATH . 'functions/imap.php');
  17 require_once(SM_PATH . 'functions/attachment_common.php');
  18 /** add sqm_baseuri()*/
  19 include_once(SM_PATH . 'functions/display_messages.php');
  20
  21 /* -------------------------------------------------------------------------- */
  22 /* MIME DECODING                                                              */
  23 /* -------------------------------------------------------------------------- */
  24
  25 /**
  26  * Get the MIME structure
  27  *
  28  * This function gets the structure of a message and stores it in the "message" class.
  29  * It will return this object for use with all relevant header information and
  30  * fully parsed into the standard "message" object format.
  31  */
  32 function mime_structure ($bodystructure, $flags=array()) {
  33
  34     /* Isolate the body structure and remove beginning and end parenthesis. */
  35     $read = trim(substr ($bodystructure, strpos(strtolower($bodystructure), 'bodystructure') + 13));
  36     $read = trim(substr ($read, 0, -1));
  37     $i = 0;
  38     $msg = Message::parseStructure($read,$i);
  39     if (!is_object($msg)) {
  40         include_once(SM_PATH . 'functions/display_messages.php');
  41         global $color, $mailbox;
  42         /* removed urldecode because $_GET is auto urldecoded ??? */
  43         displayPageHeader( $color, $mailbox );
  44         $errormessage  = _("SquirrelMail could not decode the bodystructure of the message");
  45         $errormessage .= '<br />'._("The bodystructure provided by your IMAP server:").'<br /><br />';
  46         $errormessage .= '<pre>' . htmlspecialchars($read) . '</pre>';
  47         plain_error_message( $errormessage, $color );
  48         echo '</body></html>';
  49         exit;
  50     }
  51     if (count($flags)) {
  52         foreach ($flags as $flag) {
  53             $char = strtoupper($flag{1});
  54             switch ($char) {
  55                 case 'S':
  56                     if (strtolower($flag) == '\\seen') {
  57                         $msg->is_seen = true;
  58                     }
  59                     break;
  60                 case 'A':
  61                     if (strtolower($flag) == '\\answered') {
  62                         $msg->is_answered = true;
  63                     }
  64                     break;
  65                 case 'D':
  66                     if (strtolower($flag) == '\\deleted') {
  67                         $msg->is_deleted = true;
  68                     }
  69                     break;
  70                 case 'F':
  71                     if (strtolower($flag) == '\\flagged') {
  72                         $msg->is_flagged = true;
  73                     }
  74                     break;
  75                 case 'M':
  76                     if (strtolower($flag) == '$mdnsent') {
  77                         $msg->is_mdnsent = true;
  78                     }
  79                     break;
  80                 default:
  81                     break;
  82             }
  83         }
  84     }
  85     //    listEntities($msg);
  86     return $msg;
  87 }
  88
  89
  90
  91 /* This starts the parsing of a particular structure.  It is called recursively,
  92  * so it can be passed different structures.  It returns an object of type
  93  * $message.
  94  * First, it checks to see if it is a multipart message.  If it is, then it
  95  * handles that as it sees is necessary.  If it is just a regular entity,
  96  * then it parses it and adds the necessary header information (by calling out
  97  * to mime_get_elements()
  98  */
  99
 100 function mime_fetch_body($imap_stream, $id, $ent_id=1, $fetch_size=0) {
 101     /* Do a bit of error correction.  If we couldn't find the entity id, just guess
 102      * that it is the first one.  That is usually the case anyway.
 103      */
 104
 105     if (!$ent_id) {
 106         $cmd = "FETCH $id BODY[]";
 107     } else {
 108         $cmd = "FETCH $id BODY[$ent_id]";
 109     }
 110
 111     if ($fetch_size!=0) $cmd .= "<0.$fetch_size>";
 112
 113     $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, TRUE);
 114     do {
 115         $topline = trim(array_shift($data));
 116     } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
 117
 118     $wholemessage = implode('', $data);
 119     if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
 120         $ret = substr($wholemessage, 0, $regs[1]);
 121         /* There is some information in the content info header that could be important
 122          * in order to parse html messages. Let's get them here.
 123          */
 124 //        if ($ret{0} == '<') {
 125 //            $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, TRUE);
 126 //        }
 127     } else if (ereg('"([^"]*)"', $topline, $regs)) {
 128         $ret = $regs[1];
 129     } else {
 130         global $where, $what, $mailbox, $passed_id, $startMessage;
 131         $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
 132         if (isset($where) && isset($what)) {
 133             $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
 134         } else {
 135             $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 136         }
 137         $par .= '&amp;response=' . urlencode($response) .
 138             '&amp;message='  . urlencode($message)  .
 139             '&amp;topline='  . urlencode($topline);
 140
 141         echo '<tt><br />' .
 142             '<table width="80%"><tr>' .
 143             '<tr><td colspan="2">' .
 144             _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
 145             '</td></tr>' .
 146             '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
 147             '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
 148             '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
 149             '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
 150             "</table><br /></tt></font><hr />";
 151
 152         $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, TRUE);
 153         array_shift($data);
 154         $wholemessage = implode('', $data);
 155
 156         $ret = $wholemessage;
 157     }
 158     return $ret;
 159 }
 160
 161 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding, $rStream='php://stdout') {
 162
 163     /* Don't kill the connection if the browser is over a dialup
 164      * and it would take over 30 seconds to download it.
 165      * Don't call set_time_limit in safe mode.
 166      */
 167
 168     if (!ini_get('safe_mode')) {
 169         set_time_limit(0);
 170     }
 171     /* in case of base64 encoded attachments, do not buffer them.
 172        Instead, echo the decoded attachment directly to screen */
 173     if (strtolower($encoding) == 'base64') {
 174         if (!$ent_id) {
 175             $query = "FETCH $id BODY[]";
 176         } else {
 177             $query = "FETCH $id BODY[$ent_id]";
 178         }
 179         sqimap_run_command($imap_stream,$query,true,$response,$message,TRUE,'sqimap_base64_decode',$rStream,true);
 180     } else {
 181         $body = mime_fetch_body ($imap_stream, $id, $ent_id);
 182         if (is_resource($rStream)) {
 183             fputs($rStream,decodeBody($body,$encoding));
 184         } else {
 185             echo decodeBody($body, $encoding);
 186         }
 187     }
 188
 189     /*
 190        TODO, use the same method for quoted printable.
 191        However, I assume that quoted printable attachments aren't that large
 192        so the performancegain / memory usage drop will be minimal.
 193        If we decide to add that then we need to adapt sqimap_fread because
 194        we need to split te result on \n and fread doesn't stop at \n. That
 195        means we also should provide $results from sqimap_fread (by ref) to
 196        te function and set $no_return to false. The $filter function for
 197        quoted printable should handle unsetting of $results.
 198      */
 199     /*
 200        TODO 2: find out how we write to the output stream php://stdout. fwrite
 201        doesn't work because 'php://stdout isn't a stream.
 202      */
 203
 204     return;
 205 }
 206
 207 /* -[ END MIME DECODING ]----------------------------------------------------------- */
 208
 209 /* This is here for debugging purposes.  It will print out a list
 210  * of all the entity IDs that are in the $message object.
 211  */
 212 function listEntities ($message) {
 213     if ($message) {
 214         echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br />';
 215         for ($i = 0; isset($message->entities[$i]); $i++) {
 216             echo "$i : ";
 217             $msg = listEntities($message->entities[$i]);
 218
 219             if ($msg) {
 220                 echo "return: ";
 221                 return $msg;
 222             }
 223         }
 224     }
 225 }
 226
 227 function getPriorityStr($priority) {
 228     $priority_level = substr($priority,0,1);
 229
 230     switch($priority_level) {
 231         /* Check for a higher then normal priority. */
 232         case '1':
 233         case '2':
 234             $priority_string = _("High");
 235             break;
 236
 237         /* Check for a lower then normal priority. */
 238         case '4':
 239         case '5':
 240             $priority_string = _("Low");
 241             break;
 242
 243         /* Check for a normal priority. */
 244         case '3':
 245         default:
 246             $priority_level = '3';
 247             $priority_string = _("Normal");
 248             break;
 249
 250     }
 251     return $priority_string;
 252 }
 253
 254 /* returns a $message object for a particular entity id */
 255 function getEntity ($message, $ent_id) {
 256     return $message->getEntity($ent_id);
 257 }
 258
 259 /* translateText
 260  * Extracted from strings.php 23/03/2002
 261  */
 262
 263 function translateText(&$body, $wrap_at, $charset) {
 264     global $where, $what;   /* from searching */
 265     global $color;          /* color theme */
 266
 267     require_once(SM_PATH . 'functions/url_parser.php');
 268
 269     $body_ary = explode("\n", $body);
 270     for ($i=0; $i < count($body_ary); $i++) {
 271         $line = $body_ary[$i];
 272         if (strlen($line) - 2 >= $wrap_at) {
 273             sqWordWrap($line, $wrap_at, $charset);
 274         }
 275         $line = charset_decode($charset, $line);
 276         $line = str_replace("\t", '        ', $line);
 277
 278         parseUrl ($line);
 279
 280         $quotes = 0;
 281         $pos = 0;
 282         $j = strlen($line);
 283
 284         while ($pos < $j) {
 285             if ($line[$pos] == ' ') {
 286                 $pos++;
 287             } else if (strpos($line, '&gt;', $pos) === $pos) {
 288                 $pos += 4;
 289                 $quotes++;
 290             } else {
 291                 break;
 292             }
 293         }
 294
 295         if ($quotes % 2) {
 296             if (!isset($color[13])) {
 297                 $color[13] = '#800000';
 298             }
 299             $line = '<font color="' . $color[13] . '">' . $line . '</font>';
 300         } elseif ($quotes) {
 301             if (!isset($color[14])) {
 302                 $color[14] = '#FF0000';
 303             }
 304             $line = '<font color="' . $color[14] . '">' . $line . '</font>';
 305         }
 306
 307         $body_ary[$i] = $line;
 308     }
 309     $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
 310 }
 311
 312 /**
 313  * This returns a parsed string called $body. That string can then
 314  * be displayed as the actual message in the HTML. It contains
 315  * everything needed, including HTML Tags, Attachments at the
 316  * bottom, etc.
 317  *
 318  * Since 1.2.0 function uses message_body hook.
 319  * Till 1.3.0 function included output of formatAttachments().
 320  *
 321  * @param resource $imap_stream imap connection resource
 322  * @param object $message squirrelmail message object
 323  * @param array $color squirrelmail color theme array
 324  * @param integer $wrap_at number of characters per line
 325  * @param string $ent_num (since 1.3.0) message part id
 326  * @param integer $id (since 1.3.0) message id
 327  * @param string $mailbox (since 1.3.0) imap folder name
 328  * @param boolean $clean (since 1.5.1) Do not output stuff that's irrelevant for the printable version.
 329  * @return string html formated message text
 330  */
 331 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX', $clean=FALSE) {
 332     /* This if statement checks for the entity to show as the
 333      * primary message. To add more of them, just put them in the
 334      * order that is their priority.
 335      */
 336     global $startMessage, $languages, $squirrelmail_language,
 337            $show_html_default, $sort, $has_unsafe_images, $passed_ent_id, $use_iframe,$iframe_height;
 338
 339     // workaround for not updated config.php
 340     if (! isset($use_iframe)) $use_iframe = false;
 341
 342     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
 343         $view_unsafe_images = false;
 344     }
 345
 346     $body = '';
 347     $urlmailbox = urlencode($mailbox);
 348     $body_message = getEntity($message, $ent_num);
 349     if (($body_message->header->type0 == 'text') ||
 350             ($body_message->header->type0 == 'rfc822')) {
 351         $body = mime_fetch_body ($imap_stream, $id, $ent_num);
 352         $body = decodeBody($body, $body_message->header->encoding);
 353
 354         if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 355                 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
 356             if (mb_detect_encoding($body) != 'ASCII') {
 357                 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
 358             }
 359         }
 360         $hookResults = do_hook("message_body", $body);
 361         $body = $hookResults[1];
 362
 363         /* If there are other types that shouldn't be formatted, add
 364          * them here.
 365          */
 366
 367         if ($body_message->header->type1 == 'html') {
 368             if ($show_html_default <> 1) {
 369                 $entity_conv = array('&nbsp;' => ' ',
 370                                      '<p>'    => "\n",
 371                                      '<P>'    => "\n",
 372                                      '<br>'   => "\n",
 373                                      '<BR>'   => "\n",
 374                                      '<br />' => "\n",
 375                                      '<BR />' => "\n",
 376                                      '&gt;'   => '>',
 377                                      '&lt;'   => '<');
 378                 $body = strtr($body, $entity_conv);
 379                 $body = strip_tags($body);
 380                 $body = trim($body);
 381                 translateText($body, $wrap_at,
 382                         $body_message->header->getParameter('charset'));
 383             } elseif ($use_iframe && ! $clean) {
 384                 // $clean is used to remove iframe in printable view.
 385
 386                 /**
 387                  * If we don't add html message between iframe tags,
 388                  * we must detect unsafe images and modify $has_unsafe_images.
 389                  */
 390                 $html_body = magicHTML($body, $id, $message, $mailbox);
 391                 // Convert character set in order to display html mails in different character set
 392                 $html_body = charset_decode($body_message->header->getParameter('charset'),$html_body,false,true);
 393
 394                 // creating iframe url
 395                 $iframeurl=sqm_baseuri().'src/view_html.php?'
 396                     . 'mailbox=' . $urlmailbox
 397                     . '&amp;passed_id=' . $id
 398                     . '&amp;ent_id=' . $ent_num
 399                     . '&amp;view_unsafe_images=' . (int) $view_unsafe_images;
 400
 401                 // adding warning message
 402                 $body = html_tag('div',_("Viewing HTML formatted email"),'center');
 403
 404                 /**
 405                  * height can't be set to 100%, because it does not work as expected when
 406                  * iframe is inside the table. Browsers do not create full height objects
 407                  * even when iframe is not nested. Maybe there is some way to get full size
 408                  * with CSS. Tested in firefox 1.02 and opera 7.53
 409                  *
 410                  * width="100%" does not work as expected, when table width is not set (automatic)
 411                  *
 412                  * tokul: I think <iframe> are safer sandbox than <object>. Objects might
 413                  * need special handling for IE and IE6SP2.
 414                  */
 415                 $body.= "<div><iframe name=\"message_frame\" width=\"100%\" height=\"$iframe_height\" src=\"$iframeurl\""
 416                     .' frameborder="1" marginwidth="0" marginheight="0" scrolling="auto">' . "\n";
 417
 418                 // Message for browsers without iframe support
 419                 //$body.= _("Your browser does not support inline frames.
 420                 // You can view HTML formated message by following below link.");
 421                 //$body.= "<br /><a href=\"$iframeurl\">"._("View HTML Message")."</a>";
 422
 423                 // if browser can't render iframe, it renders html message.
 424                 $body.= $html_body;
 425
 426                 // close iframe
 427                 $body.="</iframe></div>\n";
 428             } else {
 429                 // old way of html rendering
 430                 $body = magicHTML($body, $id, $message, $mailbox);
 431                 /**
 432                  * convert character set. charset_decode does not remove html special chars
 433                  * applied by magicHTML functions and does not sanitize them second time if
 434                  * fourth argument is true.
 435                  */
 436                 $body = charset_decode($body_message->header->getParameter('charset'),$body,false,true);
 437             }
 438         } else {
 439             translateText($body, $wrap_at,
 440                     $body_message->header->getParameter('charset'));
 441         }
 442
 443         // if this is the clean display (i.e. printer friendly), stop here.
 444         if ( $clean ) {
 445             return $body;
 446         }
 447
 448         $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
 449             '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
 450             '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 451         if (isset($passed_ent_id)) {
 452             $link .= '&amp;passed_ent_id='.$passed_ent_id;
 453         }
 454         $body .= '<center><small><a href="download.php?absolute_dl=true&amp;' .
 455             $link . '">' . _("Download this as a file") .  '</a>';
 456         if ($view_unsafe_images) {
 457             $text = _("Hide Unsafe Images");
 458         } else {
 459             if (isset($has_unsafe_images) && $has_unsafe_images) {
 460                 $link .= '&amp;view_unsafe_images=1';
 461                 $text = _("View Unsafe Images");
 462             } else {
 463                 $text = '';
 464             }
 465         }
 466         if($text != '') {
 467             $body .= '&nbsp;|&nbsp;<a href="read_body.php?' . $link . '">' . $text . '</a>';
 468         }
 469         $body .= '</small></center><br />' . "\n";
 470     }
 471     return $body;
 472 }
 473
 474 /**
 475  * Displays attachment links and information
 476  *
 477  * Since 1.3.0 function is not included in formatBody() call.
 478  *
 479  * Since 1.0.2 uses attachment $type0/$type1 hook.
 480  * Since 1.2.5 uses attachment $type0/* hook.
 481  * Since 1.5.0 uses attachments_bottom hook.
 482  *
 483  * @param object $message SquirrelMail message object
 484  * @param array $exclude_id message parts that are not attachments.
 485  * @param string $mailbox mailbox name
 486  * @param integer $id message id
 487  * @return string html formated attachment information.
 488  */
 489 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 490     global $where, $what, $startMessage, $color, $passed_ent_id;
 491
 492     $att_ar = $message->getAttachments($exclude_id);
 493
 494     if (!count($att_ar)) return '';
 495
 496     $attachments = '';
 497
 498     $urlMailbox = urlencode($mailbox);
 499
 500     foreach ($att_ar as $att) {
 501         $ent = $att->entity_id;
 502         $header = $att->header;
 503         $type0 = strtolower($header->type0);
 504         $type1 = strtolower($header->type1);
 505         $name = '';
 506         $links['download link']['text'] = _("Download");
 507         $links['download link']['href'] = sqm_baseuri() .
 508             "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 509         if ($type0 =='message' && $type1 == 'rfc822') {
 510             $default_page = sqm_baseuri() . 'src/read_body.php';
 511             $rfc822_header = $att->rfc822_header;
 512             $filename = $rfc822_header->subject;
 513             if (trim( $filename ) == '') {
 514                 $filename = 'untitled-[' . $ent . ']' ;
 515             }
 516             $from_o = $rfc822_header->from;
 517             if (is_object($from_o)) {
 518                 $from_name = decodeHeader($from_o->getAddress(false));
 519             } else {
 520                 $from_name = _("Unknown sender");
 521             }
 522             $description = $from_name;
 523         } else {
 524             $default_page = sqm_baseuri() . 'src/download.php';
 525             $filename = $att->getFilename();
 526             if ($header->description) {
 527                 $description = decodeHeader($header->description);
 528             } else {
 529                 $description = '';
 530             }
 531         }
 532
 533         $display_filename = $filename;
 534         if (isset($passed_ent_id)) {
 535             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 536         } else {
 537             $passed_ent_id_link = '';
 538         }
 539         $defaultlink = $default_page . "?startMessage=$startMessage"
 540             . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 541             . '&amp;ent_id='.$ent.$passed_ent_id_link;
 542         if ($where && $what) {
 543             $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 544         }
 545
 546         /* This executes the attachment hook with a specific MIME-type.
 547          * If that doesn't have results, it tries if there's a rule
 548          * for a more generic type.
 549          */
 550         $hookresults = do_hook("attachment $type0/$type1", $links,
 551                 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 552                 $display_filename, $where, $what);
 553         if(count($hookresults[1]) <= 1) {
 554             $hookresults = do_hook("attachment $type0/*", $links,
 555                     $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 556                     $display_filename, $where, $what);
 557         }
 558
 559         $links = $hookresults[1];
 560         $defaultlink = $hookresults[6];
 561
 562         $attachments .= '<tr><td>' .
 563             '<a href="'.$defaultlink.'">'.decodeHeader($display_filename).'</a>&nbsp;</td>' .
 564             '<td><small><b>' . show_readable_size($header->size) .
 565             '</b>&nbsp;&nbsp;</small></td>' .
 566             '<td><small>[ '.htmlspecialchars($type0).'/'.htmlspecialchars($type1).' ]&nbsp;</small></td>' .
 567             '<td><small>';
 568         $attachments .= '<b>' . $description . '</b>';
 569         $attachments .= '</small></td><td><small>&nbsp;';
 570
 571         $skipspaces = 1;
 572         foreach ($links as $val) {
 573             if ($skipspaces) {
 574                 $skipspaces = 0;
 575             } else {
 576                 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
 577             }
 578             $attachments .= '<a href="' . $val['href'] . '">'
 579                 . (isset($val['text']) && !empty($val['text']) ? $val['text'] : '')
 580                 . (isset($val['extra']) && !empty($val['extra']) ? $val['extra'] : '')
 581                 . '</a>';
 582         }
 583         unset($links);
 584         $attachments .= "</td></tr>\n";
 585     }
 586     $attachmentadd = do_hook_function('attachments_bottom',$attachments);
 587     if ($attachmentadd != '')
 588         $attachments = $attachmentadd;
 589     return $attachments;
 590 }
 591
 592 function sqimap_base64_decode(&$string) {
 593
 594     // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
 595     // fly decoding (to reduce memory usage) you have to check if the
 596     // data has incomplete pairs
 597
 598     // Remove the noise in order to check if the 4 bytes pairs are complete
 599     $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
 600
 601     $sStringRem = '';
 602     $iMod = strlen($string) % 4;
 603     if ($iMod) {
 604         $sStringRem = substr($string,-$iMod);
 605         // Check if $sStringRem contains padding characters
 606         if (substr($sStringRem,-1) != '=') {
 607             $string = substr($string,0,-$iMod);
 608         } else {
 609             $sStringRem = '';
 610         }
 611     }
 612     $string = base64_decode($string);
 613     return $sStringRem;
 614 }
 615
 616
 617 /* This function decodes the body depending on the encoding type. */
 618 function decodeBody($body, $encoding) {
 619     global $show_html_default;
 620
 621     $body = str_replace("\r\n", "\n", $body);
 622     $encoding = strtolower($encoding);
 623
 624     $encoding_handler = do_hook_function('decode_body', $encoding);
 625
 626
 627     // plugins get first shot at decoding the body
 628     //
 629     if (!empty($encoding_handler) && function_exists($encoding_handler)) {
 630         $body = $encoding_handler('decode', $body);
 631
 632     } else if ($encoding == 'quoted-printable' ||
 633             $encoding == 'quoted_printable') {
 634         $body = quoted_printable_decode($body);
 635
 636         while (ereg("=\n", $body)) {
 637             $body = ereg_replace ("=\n", '', $body);
 638         }
 639
 640     } else if ($encoding == 'base64') {
 641         $body = base64_decode($body);
 642     }
 643
 644     // All other encodings are returned raw.
 645     return $body;
 646 }
 647
 648 /**
 649  * Decodes headers
 650  *
 651  * This functions decode strings that is encoded according to
 652  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 653  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 654  *
 655  * @param string $string header string that has to be made readable
 656  * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
 657  * @param boolean $htmlsave preserve spaces and sanitize html special characters. defaults to true
 658  * @param boolean $decide decide if string can be utfencoded. defaults to false
 659  * @return string decoded header string
 660  */
 661 function decodeHeader ($string, $utfencode=true,$htmlsave=true,$decide=false) {
 662     global $languages, $squirrelmail_language,$default_charset;
 663     if (is_array($string)) {
 664         $string = implode("\n", $string);
 665     }
 666
 667     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 668             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
 669         $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
 670         // Do we need to return at this point?
 671         // return $string;
 672     }
 673     $i = 0;
 674     $iLastMatch = -2;
 675     $encoded = true;
 676
 677     $aString = explode(' ',$string);
 678     $ret = '';
 679     foreach ($aString as $chunk) {
 680         if ($encoded && $chunk === '') {
 681             continue;
 682         } elseif ($chunk === '') {
 683             $ret .= ' ';
 684             continue;
 685         }
 686         $encoded = false;
 687         /* if encoded words are not separated by a linear-space-white we still catch them */
 688         $j = $i-1;
 689
 690         while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 691             /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
 692             if ($iLastMatch !== $j) {
 693                 if ($htmlsave) {
 694                     $ret .= '&#32;';
 695                 } else {
 696                     $ret .= ' ';
 697                 }
 698             }
 699             $iLastMatch = $i;
 700             $j = $i;
 701             if ($htmlsave) {
 702                 $ret .= htmlspecialchars($res[1]);
 703             } else {
 704                 $ret .= $res[1];
 705             }
 706             $encoding = ucfirst($res[3]);
 707
 708             /* decide about valid decoding */
 709             if ($decide && is_conversion_safe($res[2])) {
 710                 $utfencode=true;
 711                 $can_be_encoded=true;
 712             } else {
 713                 $can_be_encoded=false;
 714             }
 715             switch ($encoding)
 716             {
 717                 case 'B':
 718                     $replace = base64_decode($res[4]);
 719                     if ($utfencode) {
 720                         if ($can_be_encoded) {
 721                             /* convert string to different charset,
 722                              * if functions asks for it (usually in compose)
 723                              */
 724                             $ret .= charset_convert($res[2],$replace,$default_charset);
 725                         } else {
 726                             // convert string to html codes in order to display it
 727                             $ret .= charset_decode($res[2],$replace);
 728                         }
 729                     } else {
 730                         if ($htmlsave) {
 731                             $replace = htmlspecialchars($replace);
 732                         }
 733                         $ret.= $replace;
 734                     }
 735                     break;
 736                 case 'Q':
 737                     $replace = str_replace('_', ' ', $res[4]);
 738                     $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 739                             $replace);
 740                     if ($utfencode) {
 741                         if ($can_be_encoded) {
 742                             /* convert string to different charset,
 743                              * if functions asks for it (usually in compose)
 744                              */
 745                             $replace = charset_convert($res[2], $replace,$default_charset);
 746                         } else {
 747                             // convert string to html codes in order to display it
 748                             $replace = charset_decode($res[2], $replace);
 749                         }
 750                     } else {
 751                         if ($htmlsave) {
 752                             $replace = htmlspecialchars($replace);
 753                         }
 754                     }
 755                     $ret .= $replace;
 756                     break;
 757                 default:
 758                     break;
 759             }
 760             $chunk = $res[5];
 761             $encoded = true;
 762         }
 763         if (!$encoded) {
 764             if ($htmlsave) {
 765                 $ret .= '&#32;';
 766             } else {
 767                 $ret .= ' ';
 768             }
 769         }
 770
 771         if (!$encoded && $htmlsave) {
 772             $ret .= htmlspecialchars($chunk);
 773         } else {
 774             $ret .= $chunk;
 775         }
 776         ++$i;
 777     }
 778     /* remove the first added space */
 779     if ($ret) {
 780         if ($htmlsave) {
 781             $ret = substr($ret,5);
 782         } else {
 783             $ret = substr($ret,1);
 784         }
 785     }
 786
 787     return $ret;
 788 }
 789
 790 /**
 791  * Encodes header
 792  *
 793  * Function uses XTRA_CODE _encodeheader function, if such function exists.
 794  *
 795  * Function uses Q encoding by default and encodes a string according to RFC
 796  * 1522 for use in headers if it contains 8-bit characters or anything that
 797  * looks like it should be encoded.
 798  *
 799  * Function switches to B encoding and encodeHeaderBase64() function, if
 800  * string is 8bit and multibyte character set supported by mbstring extension
 801  * is used. It can cause E_USER_NOTICE errors, if interface is used with
 802  * multibyte character set unsupported by mbstring extension.
 803  *
 804  * @param string $string header string, that has to be encoded
 805  * @return string quoted-printable encoded string
 806  * @todo make $mb_charsets system wide constant
 807  */
 808 function encodeHeader ($string) {
 809     global $default_charset, $languages, $squirrelmail_language;
 810
 811     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 812             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
 813         return  call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
 814     }
 815
 816     // Use B encoding for multibyte charsets
 817     $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
 818     if (in_array($default_charset,$mb_charsets) &&
 819         in_array($default_charset,sq_mb_list_encodings()) &&
 820         sq_is8bit($string)) {
 821         return encodeHeaderBase64($string,$default_charset);
 822     } elseif (in_array($default_charset,$mb_charsets) &&
 823               sq_is8bit($string) &&
 824               ! in_array($default_charset,sq_mb_list_encodings())) {
 825         // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
 826         // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
 827     }
 828
 829     // Encode only if the string contains 8-bit characters or =?
 830     $j = strlen($string);
 831     $max_l = 75 - strlen($default_charset) - 7;
 832     $aRet = array();
 833     $ret = '';
 834     $iEncStart = $enc_init = false;
 835     $cur_l = $iOffset = 0;
 836     for($i = 0; $i < $j; ++$i) {
 837         switch($string{$i})
 838         {
 839             case '=':
 840             case '<':
 841             case '>':
 842             case ',':
 843             case '?':
 844             case '_':
 845                 if ($iEncStart === false) {
 846                     $iEncStart = $i;
 847                 }
 848                 $cur_l+=3;
 849                 if ($cur_l > ($max_l-2)) {
 850                     /* if there is an stringpart that doesn't need encoding, add it */
 851                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 852                     $aRet[] = "=?$default_charset?Q?$ret?=";
 853                     $iOffset = $i;
 854                     $cur_l = 0;
 855                     $ret = '';
 856                     $iEncStart = false;
 857                 } else {
 858                     $ret .= sprintf("=%02X",ord($string{$i}));
 859                 }
 860                 break;
 861             case '(':
 862             case ')':
 863                 if ($iEncStart !== false) {
 864                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 865                     $aRet[] = "=?$default_charset?Q?$ret?=";
 866                     $iOffset = $i;
 867                     $cur_l = 0;
 868                     $ret = '';
 869                     $iEncStart = false;
 870                 }
 871                 break;
 872             case ' ':
 873                 if ($iEncStart !== false) {
 874                     $cur_l++;
 875                     if ($cur_l > $max_l) {
 876                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 877                         $aRet[] = "=?$default_charset?Q?$ret?=";
 878                         $iOffset = $i;
 879                         $cur_l = 0;
 880                         $ret = '';
 881                         $iEncStart = false;
 882                     } else {
 883                         $ret .= '_';
 884                     }
 885                 }
 886                 break;
 887             default:
 888                 $k = ord($string{$i});
 889                 if ($k > 126) {
 890                     if ($iEncStart === false) {
 891                         // do not start encoding in the middle of a string, also take the rest of the word.
 892                         $sLeadString = substr($string,0,$i);
 893                         $aLeadString = explode(' ',$sLeadString);
 894                         $sToBeEncoded = array_pop($aLeadString);
 895                         $iEncStart = $i - strlen($sToBeEncoded);
 896                         $ret .= $sToBeEncoded;
 897                         $cur_l += strlen($sToBeEncoded);
 898                     }
 899                     $cur_l += 3;
 900                     /* first we add the encoded string that reached it's max size */
 901                     if ($cur_l > ($max_l-2)) {
 902                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 903                         $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
 904                         $cur_l = 3;
 905                         $ret = '';
 906                         $iOffset = $i;
 907                         $iEncStart = $i;
 908                     }
 909                     $enc_init = true;
 910                     $ret .= sprintf("=%02X", $k);
 911                 } else {
 912                     if ($iEncStart !== false) {
 913                         $cur_l++;
 914                         if ($cur_l > $max_l) {
 915                             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 916                             $aRet[] = "=?$default_charset?Q?$ret?=";
 917                             $iEncStart = false;
 918                             $iOffset = $i;
 919                             $cur_l = 0;
 920                             $ret = '';
 921                         } else {
 922                             $ret .= $string{$i};
 923                         }
 924                     }
 925                 }
 926                 break;
 927         }
 928     }
 929
 930     if ($enc_init) {
 931         if ($iEncStart !== false) {
 932             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 933             $aRet[] = "=?$default_charset?Q?$ret?=";
 934         } else {
 935             $aRet[] = substr($string,$iOffset);
 936         }
 937         $string = implode('',$aRet);
 938     }
 939     return $string;
 940 }
 941
 942 /**
 943  * Encodes string according to rfc2047 B encoding header formating rules
 944  *
 945  * It is recommended way to encode headers with character sets that store
 946  * symbols in more than one byte.
 947  *
 948  * Function requires mbstring support. If required mbstring functions are missing,
 949  * function returns false and sets E_USER_WARNING level error message.
 950  *
 951  * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
 952  * that mbstring functions will generate E_WARNING errors, if unsupported
 953  * character set is used. mb_encode_mimeheader function provided by php
 954  * mbstring extension is not used in order to get better control of header
 955  * encoding.
 956  *
 957  * Used php code functions - function_exists(), trigger_error(), strlen()
 958  * (is used with charset names and base64 strings). Used php mbstring
 959  * functions - mb_strlen and mb_substr.
 960  *
 961  * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
 962  * encoding), rfc 2822 (header folding)
 963  *
 964  * @param string $string header string that must be encoded
 965  * @param string $charset character set. Must be supported by mbstring extension.
 966  * Use sq_mb_list_encodings() to detect supported charsets.
 967  * @return string string encoded according to rfc2047 B encoding formating rules
 968  * @since 1.5.1
 969  * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
 970  * @todo Do we want to control max length of header?
 971  * @todo Do we want to control EOL (end-of-line) marker?
 972  * @todo Do we want to translate error message?
 973  */
 974 function encodeHeaderBase64($string,$charset) {
 975     /**
 976      * Check mbstring function requirements.
 977      */
 978     if (! function_exists('mb_strlen') ||
 979         ! function_exists('mb_substr')) {
 980         // set E_USER_WARNING
 981         trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
 982         // return false
 983         return false;
 984     }
 985
 986     // initial return array
 987     $aRet = array();
 988
 989     /**
 990      * header length = 75 symbols max (same as in encodeHeader)
 991      * remove $charset length
 992      * remove =? ? ?= (5 chars)
 993      * remove 2 more chars (\r\n ?)
 994      */
 995     $iMaxLength = 75 - strlen($charset) - 7;
 996
 997     // set first character position
 998     $iStartCharNum = 0;
 999
1000     // loop through all characters. count characters and not bytes.
1001     for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1002         // encode string from starting character to current character.
1003         $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1004
1005         // Check encoded string length
1006         if(strlen($encoded_string)>$iMaxLength) {
1007             // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1008             $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1009
1010             // set new starting character
1011             $iStartCharNum = $iCharNum-1;
1012
1013             // encode last char (in case it is last character in string)
1014             $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1015         } // if string is shorter than max length - add next character
1016     }
1017
1018     // add last encoded string to array
1019     $aRet[] = $encoded_string;
1020
1021     // set initial return string
1022     $sRet = '';
1023
1024     // loop through encoded strings
1025     foreach($aRet as $string) {
1026         // TODO: Do we want to control EOL (end-of-line) marker
1027         if ($sRet!='') $sRet.= " ";
1028
1029         // add header tags and encoded string to return string
1030         $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1031     }
1032
1033     return $sRet;
1034 }
1035
1036 /* This function trys to locate the entity_id of a specific mime element */
1037 function find_ent_id($id, $message) {
1038     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1039         if ($message->entities[$i]->header->type0 == 'multipart')  {
1040             $ret = find_ent_id($id, $message->entities[$i]);
1041         } else {
1042             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
1043 //                if (sq_check_save_extension($message->entities[$i])) {
1044                 return $message->entities[$i]->entity_id;
1045 //                }
1046             } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1047                 /**
1048                  * This is part of a fix for Outlook Express 6.x generating
1049                  * cid URLs without creating content-id headers
1050                  * @@JA - 20050207
1051                  */
1052                 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1053                     return $message->entities[$i]->entity_id;
1054                 }
1055             }
1056         }
1057     }
1058     return $ret;
1059 }
1060
1061 function sq_check_save_extension($message) {
1062     $filename = $message->getFilename();
1063     $ext = substr($filename, strrpos($filename,'.')+1);
1064     $save_extensions = array('jpg','jpeg','gif','png','bmp');
1065     return in_array($ext, $save_extensions);
1066 }
1067
1068
1069 /**
1070  ** HTMLFILTER ROUTINES
1071  */
1072
1073 /**
1074  * This function checks attribute values for entity-encoded values
1075  * and returns them translated into 8-bit strings so we can run
1076  * checks on them.
1077  *
1078  * @param  $attvalue A string to run entity check against.
1079  * @return           Nothing, modifies a reference value.
1080  */
1081 function sq_defang(&$attvalue){
1082     $me = 'sq_defang';
1083     /**
1084      * Skip this if there aren't ampersands or backslashes.
1085      */
1086     if (strpos($attvalue, '&') === false
1087         && strpos($attvalue, '\\') === false){
1088         return;
1089     }
1090     $m = false;
1091     do {
1092         $m = false;
1093         $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1094         $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1095         $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1096     } while ($m == true);
1097     $attvalue = stripslashes($attvalue);
1098 }
1099
1100 /**
1101  * Kill any tabs, newlines, or carriage returns. Our friends the
1102  * makers of the browser with 95% market value decided that it'd
1103  * be funny to make "java[tab]script" be just as good as "javascript".
1104  *
1105  * @param  attvalue  The attribute value before extraneous spaces removed.
1106  * @return attvalue  Nothing, modifies a reference value.
1107  */
1108 function sq_unspace(&$attvalue){
1109     $me = 'sq_unspace';
1110     if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1111         $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1112                                 Array('',   '',   '',   '',   ''), $attvalue);
1113     }
1114 }
1115
1116 /**
1117  * This function returns the final tag out of the tag name, an array
1118  * of attributes, and the type of the tag. This function is called by
1119  * sq_sanitize internally.
1120  *
1121  * @param  $tagname  the name of the tag.
1122  * @param  $attary   the array of attributes and their values
1123  * @param  $tagtype  The type of the tag (see in comments).
1124  * @return           a string with the final tag representation.
1125  */
1126 function sq_tagprint($tagname, $attary, $tagtype){
1127     $me = 'sq_tagprint';
1128
1129     if ($tagtype == 2){
1130         $fulltag = '</' . $tagname . '>';
1131     } else {
1132         $fulltag = '<' . $tagname;
1133         if (is_array($attary) && sizeof($attary)){
1134             $atts = Array();
1135             while (list($attname, $attvalue) = each($attary)){
1136                 array_push($atts, "$attname=$attvalue");
1137             }
1138             $fulltag .= ' ' . join(" ", $atts);
1139         }
1140         if ($tagtype == 3){
1141             $fulltag .= ' /';
1142         }
1143         $fulltag .= '>';
1144     }
1145     return $fulltag;
1146 }
1147
1148 /**
1149  * A small helper function to use with array_walk. Modifies a by-ref
1150  * value and makes it lowercase.
1151  *
1152  * @param  $val a value passed by-ref.
1153  * @return      void since it modifies a by-ref value.
1154  */
1155 function sq_casenormalize(&$val){
1156     $val = strtolower($val);
1157 }
1158
1159 /**
1160  * This function skips any whitespace from the current position within
1161  * a string and to the next non-whitespace value.
1162  *
1163  * @param  $body   the string
1164  * @param  $offset the offset within the string where we should start
1165  *                 looking for the next non-whitespace character.
1166  * @return         the location within the $body where the next
1167  *                 non-whitespace char is located.
1168  */
1169 function sq_skipspace($body, $offset){
1170     $me = 'sq_skipspace';
1171     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1172     if (sizeof($matches{1})){
1173         $count = strlen($matches{1});
1174         $offset += $count;
1175     }
1176     return $offset;
1177 }
1178
1179 /**
1180  * This function looks for the next character within a string.  It's
1181  * really just a glorified "strpos", except it catches if failures
1182  * nicely.
1183  *
1184  * @param  $body   The string to look for needle in.
1185  * @param  $offset Start looking from this position.
1186  * @param  $needle The character/string to look for.
1187  * @return         location of the next occurance of the needle, or
1188  *                 strlen($body) if needle wasn't found.
1189  */
1190 function sq_findnxstr($body, $offset, $needle){
1191     $me  = 'sq_findnxstr';
1192     $pos = strpos($body, $needle, $offset);
1193     if ($pos === FALSE){
1194         $pos = strlen($body);
1195     }
1196     return $pos;
1197 }
1198
1199 /**
1200  * This function takes a PCRE-style regexp and tries to match it
1201  * within the string.
1202  *
1203  * @param  $body   The string to look for needle in.
1204  * @param  $offset Start looking from here.
1205  * @param  $reg    A PCRE-style regex to match.
1206  * @return         Returns a false if no matches found, or an array
1207  *                 with the following members:
1208  *                 - integer with the location of the match within $body
1209  *                 - string with whatever content between offset and the match
1210  *                 - string with whatever it is we matched
1211  */
1212 function sq_findnxreg($body, $offset, $reg){
1213     $me = 'sq_findnxreg';
1214     $matches = Array();
1215     $retarr = Array();
1216     preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1217     if (!isset($matches{0}) || !$matches{0}){
1218         $retarr = false;
1219     } else {
1220         $retarr{0} = $offset + strlen($matches{1});
1221         $retarr{1} = $matches{1};
1222         $retarr{2} = $matches{2};
1223     }
1224     return $retarr;
1225 }
1226
1227 /**
1228  * This function looks for the next tag.
1229  *
1230  * @param  $body   String where to look for the next tag.
1231  * @param  $offset Start looking from here.
1232  * @return         false if no more tags exist in the body, or
1233  *                 an array with the following members:
1234  *                 - string with the name of the tag
1235  *                 - array with attributes and their values
1236  *                 - integer with tag type (1, 2, or 3)
1237  *                 - integer where the tag starts (starting "<")
1238  *                 - integer where the tag ends (ending ">")
1239  *                 first three members will be false, if the tag is invalid.
1240  */
1241 function sq_getnxtag($body, $offset){
1242     $me = 'sq_getnxtag';
1243     if ($offset > strlen($body)){
1244         return false;
1245     }
1246     $lt = sq_findnxstr($body, $offset, "<");
1247     if ($lt == strlen($body)){
1248         return false;
1249     }
1250     /**
1251      * We are here:
1252      * blah blah <tag attribute="value">
1253      * \---------^
1254      */
1255     $pos = sq_skipspace($body, $lt+1);
1256     if ($pos >= strlen($body)){
1257         return Array(false, false, false, $lt, strlen($body));
1258     }
1259     /**
1260      * There are 3 kinds of tags:
1261      * 1. Opening tag, e.g.:
1262      *    <a href="blah">
1263      * 2. Closing tag, e.g.:
1264      *    </a>
1265      * 3. XHTML-style content-less tag, e.g.:
1266      *    <img src="blah" />
1267      */
1268     $tagtype = false;
1269     switch (substr($body, $pos, 1)){
1270         case '/':
1271             $tagtype = 2;
1272             $pos++;
1273             break;
1274         case '!':
1275             /**
1276              * A comment or an SGML declaration.
1277              */
1278             if (substr($body, $pos+1, 2) == "--"){
1279                 $gt = strpos($body, "-->", $pos);
1280                 if ($gt === false){
1281                     $gt = strlen($body);
1282                 } else {
1283                     $gt += 2;
1284                 }
1285                 return Array(false, false, false, $lt, $gt);
1286             } else {
1287                 $gt = sq_findnxstr($body, $pos, ">");
1288                 return Array(false, false, false, $lt, $gt);
1289             }
1290             break;
1291         default:
1292             /**
1293              * Assume tagtype 1 for now. If it's type 3, we'll switch values
1294              * later.
1295              */
1296             $tagtype = 1;
1297             break;
1298     }
1299
1300     $tag_start = $pos;
1301     $tagname = '';
1302     /**
1303      * Look for next [\W-_], which will indicate the end of the tag name.
1304      */
1305     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1306     if ($regary == false){
1307         return Array(false, false, false, $lt, strlen($body));
1308     }
1309     list($pos, $tagname, $match) = $regary;
1310     $tagname = strtolower($tagname);
1311
1312     /**
1313      * $match can be either of these:
1314      * '>'  indicating the end of the tag entirely.
1315      * '\s' indicating the end of the tag name.
1316      * '/'  indicating that this is type-3 xhtml tag.
1317      *
1318      * Whatever else we find there indicates an invalid tag.
1319      */
1320     switch ($match){
1321         case '/':
1322             /**
1323              * This is an xhtml-style tag with a closing / at the
1324              * end, like so: <img src="blah" />. Check if it's followed
1325              * by the closing bracket. If not, then this tag is invalid
1326              */
1327             if (substr($body, $pos, 2) == "/>"){
1328                 $pos++;
1329                 $tagtype = 3;
1330             } else {
1331                 $gt = sq_findnxstr($body, $pos, ">");
1332                 $retary = Array(false, false, false, $lt, $gt);
1333                 return $retary;
1334             }
1335         case '>':
1336             return Array($tagname, false, $tagtype, $lt, $pos);
1337             break;
1338         default:
1339             /**
1340              * Check if it's whitespace
1341              */
1342             if (!preg_match('/\s/', $match)){
1343                 /**
1344                  * This is an invalid tag! Look for the next closing ">".
1345                  */
1346                 $gt = sq_findnxstr($body, $lt, ">");
1347                 return Array(false, false, false, $lt, $gt);
1348             }
1349             break;
1350     }
1351
1352     /**
1353      * At this point we're here:
1354      * <tagname  attribute='blah'>
1355      * \-------^
1356      *
1357      * At this point we loop in order to find all attributes.
1358      */
1359     $attname = '';
1360     $atttype = false;
1361     $attary = Array();
1362
1363     while ($pos <= strlen($body)){
1364         $pos = sq_skipspace($body, $pos);
1365         if ($pos == strlen($body)){
1366             /**
1367              * Non-closed tag.
1368              */
1369             return Array(false, false, false, $lt, $pos);
1370         }
1371         /**
1372          * See if we arrived at a ">" or "/>", which means that we reached
1373          * the end of the tag.
1374          */
1375         $matches = Array();
1376         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1377             /**
1378              * Yep. So we did.
1379              */
1380             $pos += strlen($matches{1});
1381             if ($matches{2} == "/>"){
1382                 $tagtype = 3;
1383                 $pos++;
1384             }
1385             return Array($tagname, $attary, $tagtype, $lt, $pos);
1386         }
1387
1388         /**
1389          * There are several types of attributes, with optional
1390          * [:space:] between members.
1391          * Type 1:
1392          *   attrname[:space:]=[:space:]'CDATA'
1393          * Type 2:
1394          *   attrname[:space:]=[:space:]"CDATA"
1395          * Type 3:
1396          *   attr[:space:]=[:space:]CDATA
1397          * Type 4:
1398          *   attrname
1399          *
1400          * We leave types 1 and 2 the same, type 3 we check for
1401          * '"' and convert to "&quot" if needed, then wrap in
1402          * double quotes. Type 4 we convert into:
1403          * attrname="yes".
1404          */
1405         $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1406         if ($regary == false){
1407             /**
1408              * Looks like body ended before the end of tag.
1409              */
1410             return Array(false, false, false, $lt, strlen($body));
1411         }
1412         list($pos, $attname, $match) = $regary;
1413         $attname = strtolower($attname);
1414         /**
1415          * We arrived at the end of attribute name. Several things possible
1416          * here:
1417          * '>'  means the end of the tag and this is attribute type 4
1418          * '/'  if followed by '>' means the same thing as above
1419          * '\s' means a lot of things -- look what it's followed by.
1420          *      anything else means the attribute is invalid.
1421          */
1422         switch($match){
1423             case '/':
1424                 /**
1425                  * This is an xhtml-style tag with a closing / at the
1426                  * end, like so: <img src="blah" />. Check if it's followed
1427                  * by the closing bracket. If not, then this tag is invalid
1428                  */
1429                 if (substr($body, $pos, 2) == "/>"){
1430                     $pos++;
1431                     $tagtype = 3;
1432                 } else {
1433                     $gt = sq_findnxstr($body, $pos, ">");
1434                     $retary = Array(false, false, false, $lt, $gt);
1435                     return $retary;
1436                 }
1437             case '>':
1438                 $attary{$attname} = '"yes"';
1439                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1440                 break;
1441             default:
1442                 /**
1443                  * Skip whitespace and see what we arrive at.
1444                  */
1445                 $pos = sq_skipspace($body, $pos);
1446                 $char = substr($body, $pos, 1);
1447                 /**
1448                  * Two things are valid here:
1449                  * '=' means this is attribute type 1 2 or 3.
1450                  * \w means this was attribute type 4.
1451                  * anything else we ignore and re-loop. End of tag and
1452                  * invalid stuff will be caught by our checks at the beginning
1453                  * of the loop.
1454                  */
1455                 if ($char == "="){
1456                     $pos++;
1457                     $pos = sq_skipspace($body, $pos);
1458                     /**
1459                      * Here are 3 possibilities:
1460                      * "'"  attribute type 1
1461                      * '"'  attribute type 2
1462                      * everything else is the content of tag type 3
1463                      */
1464                     $quot = substr($body, $pos, 1);
1465                     if ($quot == "'"){
1466                         $regary = sq_findnxreg($body, $pos+1, "\'");
1467                         if ($regary == false){
1468                             return Array(false, false, false, $lt, strlen($body));
1469                         }
1470                         list($pos, $attval, $match) = $regary;
1471                         $pos++;
1472                         $attary{$attname} = "'" . $attval . "'";
1473                     } else if ($quot == '"'){
1474                         $regary = sq_findnxreg($body, $pos+1, '\"');
1475                         if ($regary == false){
1476                             return Array(false, false, false, $lt, strlen($body));
1477                         }
1478                         list($pos, $attval, $match) = $regary;
1479                         $pos++;
1480                         $attary{$attname} = '"' . $attval . '"';
1481                     } else {
1482                         /**
1483                          * These are hateful. Look for \s, or >.
1484                          */
1485                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1486                         if ($regary == false){
1487                             return Array(false, false, false, $lt, strlen($body));
1488                         }
1489                         list($pos, $attval, $match) = $regary;
1490                         /**
1491                          * If it's ">" it will be caught at the top.
1492                          */
1493                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1494                         $attary{$attname} = '"' . $attval . '"';
1495                     }
1496                 } else if (preg_match("|[\w/>]|", $char)) {
1497                     /**
1498                      * That was attribute type 4.
1499                      */
1500                     $attary{$attname} = '"yes"';
1501                 } else {
1502                     /**
1503                      * An illegal character. Find next '>' and return.
1504                      */
1505                     $gt = sq_findnxstr($body, $pos, ">");
1506                     return Array(false, false, false, $lt, $gt);
1507                 }
1508                 break;
1509         }
1510     }
1511     /**
1512      * The fact that we got here indicates that the tag end was never
1513      * found. Return invalid tag indication so it gets stripped.
1514      */
1515     return Array(false, false, false, $lt, strlen($body));
1516 }
1517
1518 /**
1519  * Translates entities into literal values so they can be checked.
1520  *
1521  * @param $attvalue the by-ref value to check.
1522  * @param $regex    the regular expression to check against.
1523  * @param $hex      whether the entites are hexadecimal.
1524  * @return          True or False depending on whether there were matches.
1525  */
1526 function sq_deent(&$attvalue, $regex, $hex=false){
1527     $me = 'sq_deent';
1528     $ret_match = false;
1529     preg_match_all($regex, $attvalue, $matches);
1530     if (is_array($matches) && sizeof($matches[0]) > 0){
1531         $repl = Array();
1532         for ($i = 0; $i < sizeof($matches[0]); $i++){
1533             $numval = $matches[1][$i];
1534             if ($hex){
1535                 $numval = hexdec($numval);
1536             }
1537             $repl{$matches[0][$i]} = chr($numval);
1538         }
1539         $attvalue = strtr($attvalue, $repl);
1540         return true;
1541     } else {
1542         return false;
1543     }
1544 }
1545
1546 /**
1547  * This function runs various checks against the attributes.
1548  *
1549  * @param  $tagname         String with the name of the tag.
1550  * @param  $attary          Array with all tag attributes.
1551  * @param  $rm_attnames     See description for sq_sanitize
1552  * @param  $bad_attvals     See description for sq_sanitize
1553  * @param  $add_attr_to_tag See description for sq_sanitize
1554  * @param  $message         message object
1555  * @param  $id              message id
1556  * @return                  Array with modified attributes.
1557  */
1558 function sq_fixatts($tagname,
1559                     $attary,
1560                     $rm_attnames,
1561                     $bad_attvals,
1562                     $add_attr_to_tag,
1563                     $message,
1564                     $id,
1565                     $mailbox
1566                     ){
1567     $me = 'sq_fixatts';
1568     while (list($attname, $attvalue) = each($attary)){
1569         /**
1570          * See if this attribute should be removed.
1571          */
1572         foreach ($rm_attnames as $matchtag=>$matchattrs){
1573             if (preg_match($matchtag, $tagname)){
1574                 foreach ($matchattrs as $matchattr){
1575                     if (preg_match($matchattr, $attname)){
1576                         unset($attary{$attname});
1577                         continue;
1578                     }
1579                 }
1580             }
1581         }
1582         /**
1583          * Remove any backslashes, entities, and extraneous whitespace.
1584          */
1585         sq_defang($attvalue);
1586         sq_unspace($attvalue);
1587
1588         /**
1589          * Now let's run checks on the attvalues.
1590          * I don't expect anyone to comprehend this. If you do,
1591          * get in touch with me so I can drive to where you live and
1592          * shake your hand personally. :)
1593          */
1594         foreach ($bad_attvals as $matchtag=>$matchattrs){
1595             if (preg_match($matchtag, $tagname)){
1596                 foreach ($matchattrs as $matchattr=>$valary){
1597                     if (preg_match($matchattr, $attname)){
1598                         /**
1599                          * There are two arrays in valary.
1600                          * First is matches.
1601                          * Second one is replacements
1602                          */
1603                         list($valmatch, $valrepl) = $valary;
1604                         $newvalue =
1605                             preg_replace($valmatch, $valrepl, $attvalue);
1606                         if ($newvalue != $attvalue){
1607                             $attary{$attname} = $newvalue;
1608                         }
1609                     }
1610                 }
1611             }
1612         }
1613
1614         /**
1615          * Replace empty src tags with the blank image.  src is only used
1616          * for frames, images, and image inputs.  Doing a replace should
1617          * not affect them working as should be, however it will stop
1618          * IE from being kicked off when src for img tags are not set
1619          */
1620         if (($attname == 'src') && ($attvalue == '""')) {
1621             $attary{$attname} = '"' . SM_PATH . 'images/blank.png"';
1622         }
1623
1624         /**
1625          * Turn cid: urls into http-friendly ones.
1626          */
1627         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1628             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1629         }
1630
1631         /**
1632          * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1633          * One day MS might actually make it match something useful, for now, falling
1634          * back to using cid2http, so we can grab the blank.png.
1635          */
1636         if (preg_match("/^[\'\"]\s*outbind:\/\//si", $attvalue)) {
1637             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1638         }
1639
1640     }
1641     /**
1642      * See if we need to append any attributes to this tag.
1643      */
1644     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1645         if (preg_match($matchtag, $tagname)){
1646             $attary = array_merge($attary, $addattary);
1647         }
1648     }
1649     return $attary;
1650 }
1651
1652 /**
1653  * This function edits the style definition to make them friendly and
1654  * usable in SquirrelMail.
1655  *
1656  * @param  $message  the message object
1657  * @param  $id       the message id
1658  * @param  $content  a string with whatever is between <style> and </style>
1659  * @param  $mailbox  the message mailbox
1660  * @return           a string with edited content.
1661  */
1662 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1663     global $view_unsafe_images;
1664     $me = 'sq_fixstyle';
1665     $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1666     if ($ret == FALSE){
1667         return array(FALSE, strlen($body));
1668     }
1669     $newpos = $ret[0] + strlen($ret[2]);
1670     $content = $ret[1];
1671     /**
1672     * First look for general BODY style declaration, which would be
1673     * like so:
1674     * body {background: blah-blah}
1675     * and change it to .bodyclass so we can just assign it to a <div>
1676     */
1677     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1678     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1679     /**
1680     * Fix url('blah') declarations.
1681     */
1682     //   $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1683     //                           "url(\\1$secremoveimg\\2)", $content);
1684     // remove NUL
1685     $content = str_replace("\0", "", $content);
1686     // NB I insert NUL characters to keep to avoid an infinite loop. They are removed after the loop.
1687     while (preg_match("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si", $content, $matches)) {
1688         $sProto = strtolower($matches[1]);
1689         switch ($sProto) {
1690           /**
1691            * Fix url('https*://.*) declarations but only if $view_unsafe_images
1692            * is false.
1693            */
1694           case 'https':
1695           case 'http':
1696             if (!$view_unsafe_images){
1697                 $sExpr = "/url\s*\(\s*([\'\"])\s*$sProto*:.*?([\'\"])\s*\)/si";
1698                 $content = preg_replace($sExpr, "u\0r\0l(\\1$secremoveimg\\2)", $content);
1699             }
1700             break;
1701           /**
1702            * Fix urls that refer to cid:
1703            */
1704           case 'cid':
1705             $cidurl = 'cid:'. $matches[2];
1706             $httpurl = sq_cid2http($message, $id, $cidurl, $mailbox);
1707             $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1708                                 "u\0r\0l($httpurl)", $content);
1709             break;
1710           default:
1711             /**
1712              * replace url with protocol other then the white list
1713              * http,https and cid by an empty string.
1714              */
1715             $content = preg_replace("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si",
1716                                 "", $content);
1717             break;
1718         }
1719         break;
1720     }
1721     // remove NUL
1722     $content = str_replace("\0", "", $content);
1723
1724    /**
1725     * Remove any backslashes, entities, and extraneous whitespace.
1726     */
1727     $contentTemp = $content;
1728     sq_defang($contentTemp);
1729     sq_unspace($contentTemp);
1730
1731     /**
1732      * Fix stupid css declarations which lead to vulnerabilities
1733      * in IE.
1734      */
1735     $match   = Array('/expression/i',
1736                     '/behaviou*r/i',
1737                     '/binding/i',
1738                     '/include-source/i');
1739     $replace = Array('idiocy', 'idiocy', 'idiocy', 'idiocy');
1740     $contentNew = preg_replace($match, $replace, $contentTemp);
1741     if ($contentNew !== $contentTemp) {
1742         // insecure css declarations are used. From now on we don't care
1743         // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
1744         $content = $contentNew;
1745     }
1746     return array($content, $newpos);
1747 }
1748
1749
1750 /**
1751  * This function converts cid: url's into the ones that can be viewed in
1752  * the browser.
1753  *
1754  * @param  $message  the message object
1755  * @param  $id       the message id
1756  * @param  $cidurl   the cid: url.
1757  * @param  $mailbox  the message mailbox
1758  * @return           a string with a http-friendly url
1759  */
1760 function sq_cid2http($message, $id, $cidurl, $mailbox){
1761     /**
1762      * Get rid of quotes.
1763      */
1764     $quotchar = substr($cidurl, 0, 1);
1765     if ($quotchar == '"' || $quotchar == "'"){
1766         $cidurl = str_replace($quotchar, "", $cidurl);
1767     } else {
1768         $quotchar = '';
1769     }
1770     $cidurl = substr(trim($cidurl), 4);
1771
1772     $match_str = '/\{.*?\}\//';
1773     $str_rep = '';
1774     $cidurl = preg_replace($match_str, $str_rep, $cidurl);
1775
1776     $linkurl = find_ent_id($cidurl, $message);
1777     /* in case of non-save cid links $httpurl should be replaced by a sort of
1778        unsave link image */
1779     $httpurl = '';
1780
1781     /**
1782      * This is part of a fix for Outlook Express 6.x generating
1783      * cid URLs without creating content-id headers. These images are
1784      * not part of the multipart/related html mail. The html contains
1785      * <img src="cid:{some_id}/image_filename.ext"> references to
1786      * attached images with as goal to render them inline although
1787      * the attachment disposition property is not inline.
1788      */
1789
1790     if (empty($linkurl)) {
1791         if (preg_match('/{.*}\//', $cidurl)) {
1792             $cidurl = preg_replace('/{.*}\//','', $cidurl);
1793             if (!empty($cidurl)) {
1794                 $linkurl = find_ent_id($cidurl, $message);
1795             }
1796         }
1797     }
1798
1799     if (!empty($linkurl)) {
1800         $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1801             "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1802             '&amp;ent_id=' . $linkurl . $quotchar;
1803     } else {
1804         /**
1805          * If we couldn't generate a proper img url, drop in a blank image
1806          * instead of sending back empty, otherwise it causes unusual behaviour
1807          */
1808         $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
1809     }
1810
1811     return $httpurl;
1812 }
1813
1814 /**
1815  * This function changes the <body> tag into a <div> tag since we
1816  * can't really have a body-within-body.
1817  *
1818  * @param  $attary   an array of attributes and values of <body>
1819  * @param  $mailbox  mailbox we're currently reading (for cid2http)
1820  * @param  $message  current message (for cid2http)
1821  * @param  $id       current message id (for cid2http)
1822  * @return           a modified array of attributes to be set for <div>
1823  */
1824 function sq_body2div($attary, $mailbox, $message, $id){
1825     $me = 'sq_body2div';
1826     $divattary = Array('class' => "'bodyclass'");
1827     $text = '#000000';
1828     $has_bgc_stl = $has_txt_stl = false;
1829     $styledef = '';
1830     if (is_array($attary) && sizeof($attary) > 0){
1831         foreach ($attary as $attname=>$attvalue){
1832             $quotchar = substr($attvalue, 0, 1);
1833             $attvalue = str_replace($quotchar, "", $attvalue);
1834             switch ($attname){
1835                 case 'background':
1836                     $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1837                     $styledef .= "background-image: url('$attvalue'); ";
1838                     break;
1839                 case 'bgcolor':
1840                     $has_bgc_stl = true;
1841                     $styledef .= "background-color: $attvalue; ";
1842                     break;
1843                 case 'text':
1844                     $has_txt_stl = true;
1845                     $styledef .= "color: $attvalue; ";
1846                     break;
1847             }
1848         }
1849         // Outlook defines a white bgcolor and no text color. This can lead to
1850         // white text on a white bg with certain themes.
1851         if ($has_bgc_stl && !$has_txt_stl) {
1852             $styledef .= "color: $text; ";
1853         }
1854         if (strlen($styledef) > 0){
1855             $divattary{"style"} = "\"$styledef\"";
1856         }
1857     }
1858     return $divattary;
1859 }
1860
1861 /**
1862  * This is the main function and the one you should actually be calling.
1863  * There are several variables you should be aware of an which need
1864  * special description.
1865  *
1866  * Since the description is quite lengthy, see it here:
1867  * http://linux.duke.edu/projects/mini/htmlfilter/
1868  *
1869  * @param $body                 the string with HTML you wish to filter
1870  * @param $tag_list             see description above
1871  * @param $rm_tags_with_content see description above
1872  * @param $self_closing_tags    see description above
1873  * @param $force_tag_closing    see description above
1874  * @param $rm_attnames          see description above
1875  * @param $bad_attvals          see description above
1876  * @param $add_attr_to_tag      see description above
1877  * @param $message              message object
1878  * @param $id                   message id
1879  * @return                      sanitized html safe to show on your pages.
1880  */
1881 function sq_sanitize($body,
1882                      $tag_list,
1883                      $rm_tags_with_content,
1884                      $self_closing_tags,
1885                      $force_tag_closing,
1886                      $rm_attnames,
1887                      $bad_attvals,
1888                      $add_attr_to_tag,
1889                      $message,
1890                      $id,
1891                      $mailbox
1892                      ){
1893     $me = 'sq_sanitize';
1894     $rm_tags = array_shift($tag_list);
1895     /**
1896      * Normalize rm_tags and rm_tags_with_content.
1897      */
1898     @array_walk($tag_list, 'sq_casenormalize');
1899     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1900     @array_walk($self_closing_tags, 'sq_casenormalize');
1901     /**
1902      * See if tag_list is of tags to remove or tags to allow.
1903      * false  means remove these tags
1904      * true   means allow these tags
1905      */
1906     $curpos = 0;
1907     $open_tags = Array();
1908     $trusted = "\n<!-- begin sanitized html -->\n";
1909     $skip_content = false;
1910     /**
1911      * Take care of netscape's stupid javascript entities like
1912      * &{alert('boo')};
1913      */
1914     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1915
1916     while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1917         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1918         $free_content = substr($body, $curpos, $lt-$curpos);
1919         /**
1920          * Take care of <style>
1921          */
1922         if ($tagname == "style" && $tagtype == 1){
1923             list($free_content, $curpos) =
1924                 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
1925             if ($free_content != FALSE){
1926                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1927                 $trusted .= $free_content;
1928                 $trusted .= sq_tagprint($tagname, false, 2);
1929             }
1930             continue;
1931         }
1932         if ($skip_content == false){
1933             $trusted .= $free_content;
1934         }
1935         if ($tagname != FALSE){
1936             if ($tagtype == 2){
1937                 if ($skip_content == $tagname){
1938                     /**
1939                      * Got to the end of tag we needed to remove.
1940                      */
1941                     $tagname = false;
1942                     $skip_content = false;
1943                 } else {
1944                     if ($skip_content == false){
1945                         if ($tagname == "body"){
1946                             $tagname = "div";
1947                         }
1948                         if (isset($open_tags{$tagname}) &&
1949                                 $open_tags{$tagname} > 0){
1950                             $open_tags{$tagname}--;
1951                         } else {
1952                             $tagname = false;
1953                         }
1954                     }
1955                 }
1956             } else {
1957                 /**
1958                  * $rm_tags_with_content
1959                  */
1960                 if ($skip_content == false){
1961                     /**
1962                      * See if this is a self-closing type and change
1963                      * tagtype appropriately.
1964                      */
1965                     if ($tagtype == 1
1966                             && in_array($tagname, $self_closing_tags)){
1967                         $tagtype = 3;
1968                     }
1969                     /**
1970                      * See if we should skip this tag and any content
1971                      * inside it.
1972                      */
1973                     if ($tagtype == 1 &&
1974                             in_array($tagname, $rm_tags_with_content)){
1975                         $skip_content = $tagname;
1976                     } else {
1977                         if (($rm_tags == false
1978                                     && in_array($tagname, $tag_list)) ||
1979                                 ($rm_tags == true &&
1980                                  !in_array($tagname, $tag_list))){
1981                             $tagname = false;
1982                         } else {
1983                             /**
1984                              * Convert body into div.
1985                              */
1986                             if ($tagname == "body"){
1987                                 $tagname = "div";
1988                                 $attary = sq_body2div($attary, $mailbox,
1989                                         $message, $id);
1990                             }
1991                             if ($tagtype == 1){
1992                                 if (isset($open_tags{$tagname})){
1993                                     $open_tags{$tagname}++;
1994                                 } else {
1995                                     $open_tags{$tagname}=1;
1996                                 }
1997                             }
1998                             /**
1999                              * This is where we run other checks.
2000                              */
2001                             if (is_array($attary) && sizeof($attary) > 0){
2002                                 $attary = sq_fixatts($tagname,
2003                                                      $attary,
2004                                                      $rm_attnames,
2005                                                      $bad_attvals,
2006                                                      $add_attr_to_tag,
2007                                                      $message,
2008                                                      $id,
2009                                                      $mailbox
2010                                                      );
2011                             }
2012                         }
2013                     }
2014                 }
2015             }
2016             if ($tagname != false && $skip_content == false){
2017                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2018             }
2019         }
2020         $curpos = $gt+1;
2021     }
2022     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2023     if ($force_tag_closing == true){
2024         foreach ($open_tags as $tagname=>$opentimes){
2025             while ($opentimes > 0){
2026                 $trusted .= '</' . $tagname . '>';
2027                 $opentimes--;
2028             }
2029         }
2030         $trusted .= "\n";
2031     }
2032     $trusted .= "<!-- end sanitized html -->\n";
2033     return $trusted;
2034 }
2035
2036 /**
2037  * This is a wrapper function to call html sanitizing routines.
2038  *
2039  * @param  $body  the body of the message
2040  * @param  $id    the id of the message
2041
2042  * @param  $message
2043  * @param  $mailbox
2044  * @param  boolean $take_mailto_links When TRUE, converts mailto: links
2045  *                                    into internal SM compose links
2046  *                                    (optional; default = TRUE)
2047  * @return        a string with html safe to display in the browser.
2048  */
2049 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2050
2051     require_once(SM_PATH . 'functions/url_parser.php');  // for $MailTo_PReg_Match
2052
2053     global $attachment_common_show_images, $view_unsafe_images,
2054            $has_unsafe_images;
2055     /**
2056      * Don't display attached images in HTML mode.
2057      */
2058     $attachment_common_show_images = false;
2059     $tag_list = Array(
2060             false,
2061             "object",
2062             "meta",
2063             "html",
2064             "head",
2065             "base",
2066             "link",
2067             "frame",
2068             "iframe",
2069             "plaintext",
2070             "marquee"
2071             );
2072
2073     $rm_tags_with_content = Array(
2074             "script",
2075             "applet",
2076             "embed",
2077             "title",
2078             "frameset",
2079             "xmp",
2080             "xml"
2081             );
2082
2083     $self_closing_tags =  Array(
2084             "img",
2085             "br",
2086             "hr",
2087             "input",
2088             "outbind"
2089             );
2090
2091     $force_tag_closing = true;
2092
2093     $rm_attnames = Array(
2094             "/.*/" =>
2095             Array(
2096                 "/target/i",
2097                 "/^on.*/i",
2098                 "/^dynsrc/i",
2099                 "/^data.*/i",
2100                 "/^lowsrc.*/i"
2101                 )
2102             );
2103
2104     $secremoveimg = "../images/" . _("sec_remove_eng.png");
2105     $bad_attvals = Array(
2106             "/.*/" =>
2107             Array(
2108                 "/^src|background/i" =>
2109                 Array(
2110                     Array(
2111                         "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2112                         "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2113                         "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2114                         ),
2115                     Array(
2116                         "\\1$secremoveimg\\2",
2117                         "\\1$secremoveimg\\2",
2118                         "\\1$secremoveimg\\2",
2119                         )
2120                     ),
2121                 "/^href|action/i" =>
2122                 Array(
2123                     Array(
2124                         "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2125                         "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2126                         "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2127                         ),
2128                     Array(
2129                         "\\1#\\1",
2130                         "\\1#\\1",
2131                         "\\1#\\1"
2132                         )
2133                     ),
2134         "/^style/i" =>
2135             Array(
2136                 Array(
2137                     "/expression/i",
2138                     "/binding/i",
2139                     "/behaviou*r/i",
2140                     "/include-source/i",
2141                     "/position\s*:\s*absolute/i",
2142                     "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2143                     "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2144                     "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2145                     "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2146                     ),
2147                 Array(
2148                     "idiocy",
2149                     "idiocy",
2150                     "idiocy",
2151                     "idiocy",
2152                     "",
2153                     "url(\\1#\\1)",
2154                     "url(\\1#\\1)",
2155                     "url(\\1#\\1)",
2156                     "\\1:url(\\2#\\3)"
2157                     )
2158                 )
2159             )
2160         );
2161     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
2162         $view_unsafe_images = false;
2163     }
2164     if (!$view_unsafe_images){
2165         /**
2166          * Remove any references to http/https if view_unsafe_images set
2167          * to false.
2168          */
2169         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2170                 '/^([\'\"])\s*https*:.*([\'\"])/si');
2171         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2172                 "\\1$secremoveimg\\1");
2173         array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2174                 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2175         array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2176                 "url(\\1$secremoveimg\\1)");
2177     }
2178
2179     $add_attr_to_tag = Array(
2180             "/^a$/i" =>
2181             Array('target'=>'"_blank"',
2182                 'title'=>'"'._("This external link will open in a new window").'"'
2183                 )
2184             );
2185     $trusted = sq_sanitize($body,
2186                            $tag_list,
2187                            $rm_tags_with_content,
2188                            $self_closing_tags,
2189                            $force_tag_closing,
2190                            $rm_attnames,
2191                            $bad_attvals,
2192                            $add_attr_to_tag,
2193                            $message,
2194                            $id,
2195                            $mailbox
2196                            );
2197     if (preg_match("|$secremoveimg|i", $trusted)){
2198         $has_unsafe_images = true;
2199     }
2200
2201     // we want to parse mailto's in HTML output, change to SM compose links
2202     // this is a modified version of code from url_parser.php... but Marc is
2203     // right: we need a better filtering implementation; adding this randomly
2204     // here is not a great solution
2205     //
2206     if ($take_mailto_links) {
2207         // parseUrl($trusted);   // this even parses URLs inside of tags... too aggressive
2208         global $MailTo_PReg_Match;
2209         $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1);
2210         if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2211             foreach ($regs[0] as $i => $mailto_before) {
2212                 $mailto_params = $regs[10][$i];
2213                 // get rid of any tailing quote since we have to add send_to to the end
2214                 //
2215                 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2216                     $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2217                 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2218                     $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2219
2220                 if ($regs[1][$i]) {    //if there is an email addr before '?', we need to merge it with the params
2221                     $to = 'to=' . $regs[1][$i];
2222                     if (strpos($mailto_params, 'to=') > -1)    //already a 'to='
2223                         $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2224                     else {
2225                         if ($mailto_params)    //already some params, append to them
2226                             $mailto_params .= '&amp;' . $to;
2227                         else
2228                             $mailto_params .= '?' . $to;
2229                     }
2230                 }
2231
2232                 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2233
2234                 // we'll already have target=_blank, no need to allow comp_in_new
2235                 // here (which would be a lot more work anyway)
2236                 //
2237                 global $compose_new_win;
2238                 $temp_comp_in_new = $compose_new_win;
2239                 $compose_new_win = 0;
2240                 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2241                 $compose_new_win = $temp_comp_in_new;
2242
2243                 // remove <a href=" and anything after the next quote (we only
2244                 // need the uri, not the link HTML) in compose uri
2245                 //
2246                 $comp_uri = substr($comp_uri, 9);
2247                 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2248                 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2249             }
2250         }
2251     }
2252
2253     return $trusted;
2254 }
2255
2256 /**
2257  * function SendDownloadHeaders - send file to the browser
2258  *
2259  * Original Source: SM core src/download.php
2260  * moved here to make it available to other code, and separate
2261  * front end from back end functionality.
2262  *
2263  * @param string $type0 first half of mime type
2264  * @param string $type1 second half of mime type
2265  * @param string $filename filename to tell the browser for downloaded file
2266  * @param boolean $force whether to force the download dialog to pop
2267  * @param optional integer $filesize send the Content-Header and length to the browser
2268  * @return void
2269  */
2270 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2271     global $languages, $squirrelmail_language;
2272     $isIE = $isIE6 = 0;
2273
2274     sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2275
2276     if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2277             strstr($HTTP_USER_AGENT, 'Opera') === false) {
2278         $isIE = 1;
2279     }
2280
2281     if (strstr($HTTP_USER_AGENT, 'compatible; MSIE 6') !== false &&
2282             strstr($HTTP_USER_AGENT, 'Opera') === false) {
2283         $isIE6 = 1;
2284     }
2285
2286     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2287             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2288         $filename =
2289             call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2290     } else {
2291         $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
2292     }
2293
2294     // A Pox on Microsoft and it's Internet Explorer!
2295     //
2296     // IE has lots of bugs with file downloads.
2297     // It also has problems with SSL.  Both of these cause problems
2298     // for us in this function.
2299     //
2300     // See this article on Cache Control headers and SSL
2301     // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2302     //
2303     // The best thing you can do for IE is to upgrade to the latest
2304     // version
2305     //set all the Cache Control Headers for IE
2306     if ($isIE) {
2307         $filename=rawurlencode($filename);
2308         header ("Pragma: public");
2309         header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2310         header ("Cache-Control: post-check=0, pre-check=0", false);
2311         header ("Cache-Control: private");
2312
2313         //set the inline header for IE, we'll add the attachment header later if we need it
2314         header ("Content-Disposition: inline; filename=$filename");
2315     }
2316
2317     if (!$force) {
2318         // Try to show in browser window
2319         header ("Content-Disposition: inline; filename=\"$filename\"");
2320         header ("Content-Type: $type0/$type1; name=\"$filename\"");
2321     } else {
2322         // Try to pop up the "save as" box
2323
2324         // IE makes this hard.  It pops up 2 save boxes, or none.
2325         // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2326         // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2327         // But, according to Microsoft, it is "RFC compliant but doesn't
2328         // take into account some deviations that allowed within the
2329         // specification."  Doesn't that mean RFC non-compliant?
2330         // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2331
2332         // all browsers need the application/octet-stream header for this
2333         header ("Content-Type: application/octet-stream; name=\"$filename\"");
2334
2335         // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2336         // Do not have quotes around filename, but that applied to
2337         // "attachment"... does it apply to inline too?
2338         header ("Content-Disposition: attachment; filename=\"$filename\"");
2339
2340         if ($isIE && !$isIE6) {
2341             // This combination seems to work mostly.  IE 5.5 SP 1 has
2342             // known issues (see the Microsoft Knowledge Base)
2343
2344             // This works for most types, but doesn't work with Word files
2345             header ("Content-Type: application/download; name=\"$filename\"");
2346
2347             // These are spares, just in case.  :-)
2348             //header("Content-Type: $type0/$type1; name=\"$filename\"");
2349             //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2350             //header("Content-Type: application/octet-stream; name=\"$filename\"");
2351         } else {
2352             // another application/octet-stream forces download for Netscape
2353             header ("Content-Type: application/octet-stream; name=\"$filename\"");
2354         }
2355     }
2356
2357     //send the content-length header if the calling function provides it
2358     if ($filesize > 0) {
2359         header("Content-Length: $filesize");
2360     }
2361
2362 }  // end fn SendDownloadHeaders
2363
2364 ?>