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=\"$iframeurl\""
 416                     .' frameborder="1" marginwidth="0" marginheight="0" scrolling="auto" security="restricted">' . "\n";
 417
 418                 // Message for browsers without iframe support
 419                 //$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  * Decodes encoded message body
 618  *
 619  * This function decodes the body depending on the encoding type.
 620  * Currently quoted-printable and base64 encodings are supported.
 621  * decode_body hook was added to this function in 1.4.2/1.5.0
 622  * @param string $body encoded message body
 623  * @param string $encoding used encoding
 624  * @return string decoded string
 625  * @since 1.0
 626  */
 627 function decodeBody($body, $encoding) {
 628
 629     $body = str_replace("\r\n", "\n", $body);
 630     $encoding = strtolower($encoding);
 631
 632     $encoding_handler = do_hook_function('decode_body', $encoding);
 633
 634
 635     // plugins get first shot at decoding the body
 636     //
 637     if (!empty($encoding_handler) && function_exists($encoding_handler)) {
 638         $body = $encoding_handler('decode', $body);
 639
 640     } elseif ($encoding == 'quoted-printable' ||
 641             $encoding == 'quoted_printable') {
 642         /**
 643          * quoted_printable_decode() function is broken in older
 644          * php versions. Text with \r\n decoding was fixed only
 645          * in php 4.3.0. Minimal code requirement 4.0.4 +
 646          * str_replace("\r\n", "\n", $body); call.
 647          */
 648         $body = quoted_printable_decode($body);
 649     } elseif ($encoding == 'base64') {
 650         $body = base64_decode($body);
 651     }
 652
 653     // All other encodings are returned raw.
 654     return $body;
 655 }
 656
 657 /**
 658  * Decodes headers
 659  *
 660  * This functions decode strings that is encoded according to
 661  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 662  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 663  *
 664  * @param string $string header string that has to be made readable
 665  * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
 666  * @param boolean $htmlsave preserve spaces and sanitize html special characters. defaults to true
 667  * @param boolean $decide decide if string can be utfencoded. defaults to false
 668  * @return string decoded header string
 669  */
 670 function decodeHeader ($string, $utfencode=true,$htmlsave=true,$decide=false) {
 671     global $languages, $squirrelmail_language,$default_charset;
 672     if (is_array($string)) {
 673         $string = implode("\n", $string);
 674     }
 675
 676     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 677             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
 678         $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
 679         // Do we need to return at this point?
 680         // return $string;
 681     }
 682     $i = 0;
 683     $iLastMatch = -2;
 684     $encoded = true;
 685
 686     $aString = explode(' ',$string);
 687     $ret = '';
 688     foreach ($aString as $chunk) {
 689         if ($encoded && $chunk === '') {
 690             continue;
 691         } elseif ($chunk === '') {
 692             $ret .= ' ';
 693             continue;
 694         }
 695         $encoded = false;
 696         /* if encoded words are not separated by a linear-space-white we still catch them */
 697         $j = $i-1;
 698
 699         while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 700             /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
 701             if ($iLastMatch !== $j) {
 702                 if ($htmlsave) {
 703                     $ret .= '&#32;';
 704                 } else {
 705                     $ret .= ' ';
 706                 }
 707             }
 708             $iLastMatch = $i;
 709             $j = $i;
 710             if ($htmlsave) {
 711                 $ret .= htmlspecialchars($res[1]);
 712             } else {
 713                 $ret .= $res[1];
 714             }
 715             $encoding = ucfirst($res[3]);
 716
 717             /* decide about valid decoding */
 718             if ($decide && is_conversion_safe($res[2])) {
 719                 $utfencode=true;
 720                 $can_be_encoded=true;
 721             } else {
 722                 $can_be_encoded=false;
 723             }
 724             switch ($encoding)
 725             {
 726                 case 'B':
 727                     $replace = base64_decode($res[4]);
 728                     if ($utfencode) {
 729                         if ($can_be_encoded) {
 730                             /* convert string to different charset,
 731                              * if functions asks for it (usually in compose)
 732                              */
 733                             $ret .= charset_convert($res[2],$replace,$default_charset);
 734                         } else {
 735                             // convert string to html codes in order to display it
 736                             $ret .= charset_decode($res[2],$replace);
 737                         }
 738                     } else {
 739                         if ($htmlsave) {
 740                             $replace = htmlspecialchars($replace);
 741                         }
 742                         $ret.= $replace;
 743                     }
 744                     break;
 745                 case 'Q':
 746                     $replace = str_replace('_', ' ', $res[4]);
 747                     $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 748                             $replace);
 749                     if ($utfencode) {
 750                         if ($can_be_encoded) {
 751                             /* convert string to different charset,
 752                              * if functions asks for it (usually in compose)
 753                              */
 754                             $replace = charset_convert($res[2], $replace,$default_charset);
 755                         } else {
 756                             // convert string to html codes in order to display it
 757                             $replace = charset_decode($res[2], $replace);
 758                         }
 759                     } else {
 760                         if ($htmlsave) {
 761                             $replace = htmlspecialchars($replace);
 762                         }
 763                     }
 764                     $ret .= $replace;
 765                     break;
 766                 default:
 767                     break;
 768             }
 769             $chunk = $res[5];
 770             $encoded = true;
 771         }
 772         if (!$encoded) {
 773             if ($htmlsave) {
 774                 $ret .= '&#32;';
 775             } else {
 776                 $ret .= ' ';
 777             }
 778         }
 779
 780         if (!$encoded && $htmlsave) {
 781             $ret .= htmlspecialchars($chunk);
 782         } else {
 783             $ret .= $chunk;
 784         }
 785         ++$i;
 786     }
 787     /* remove the first added space */
 788     if ($ret) {
 789         if ($htmlsave) {
 790             $ret = substr($ret,5);
 791         } else {
 792             $ret = substr($ret,1);
 793         }
 794     }
 795
 796     return $ret;
 797 }
 798
 799 /**
 800  * Encodes header
 801  *
 802  * Function uses XTRA_CODE _encodeheader function, if such function exists.
 803  *
 804  * Function uses Q encoding by default and encodes a string according to RFC
 805  * 1522 for use in headers if it contains 8-bit characters or anything that
 806  * looks like it should be encoded.
 807  *
 808  * Function switches to B encoding and encodeHeaderBase64() function, if
 809  * string is 8bit and multibyte character set supported by mbstring extension
 810  * is used. It can cause E_USER_NOTICE errors, if interface is used with
 811  * multibyte character set unsupported by mbstring extension.
 812  *
 813  * @param string $string header string, that has to be encoded
 814  * @return string quoted-printable encoded string
 815  * @todo make $mb_charsets system wide constant
 816  */
 817 function encodeHeader ($string) {
 818     global $default_charset, $languages, $squirrelmail_language;
 819
 820     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 821             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
 822         return  call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
 823     }
 824
 825     // Use B encoding for multibyte charsets
 826     $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
 827     if (in_array($default_charset,$mb_charsets) &&
 828         in_array($default_charset,sq_mb_list_encodings()) &&
 829         sq_is8bit($string)) {
 830         return encodeHeaderBase64($string,$default_charset);
 831     } elseif (in_array($default_charset,$mb_charsets) &&
 832               sq_is8bit($string) &&
 833               ! in_array($default_charset,sq_mb_list_encodings())) {
 834         // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
 835         // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
 836     }
 837
 838     // Encode only if the string contains 8-bit characters or =?
 839     $j = strlen($string);
 840     $max_l = 75 - strlen($default_charset) - 7;
 841     $aRet = array();
 842     $ret = '';
 843     $iEncStart = $enc_init = false;
 844     $cur_l = $iOffset = 0;
 845     for($i = 0; $i < $j; ++$i) {
 846         switch($string{$i})
 847         {
 848             case '=':
 849             case '<':
 850             case '>':
 851             case ',':
 852             case '?':
 853             case '_':
 854                 if ($iEncStart === false) {
 855                     $iEncStart = $i;
 856                 }
 857                 $cur_l+=3;
 858                 if ($cur_l > ($max_l-2)) {
 859                     /* if there is an stringpart that doesn't need encoding, add it */
 860                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 861                     $aRet[] = "=?$default_charset?Q?$ret?=";
 862                     $iOffset = $i;
 863                     $cur_l = 0;
 864                     $ret = '';
 865                     $iEncStart = false;
 866                 } else {
 867                     $ret .= sprintf("=%02X",ord($string{$i}));
 868                 }
 869                 break;
 870             case '(':
 871             case ')':
 872                 if ($iEncStart !== false) {
 873                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 874                     $aRet[] = "=?$default_charset?Q?$ret?=";
 875                     $iOffset = $i;
 876                     $cur_l = 0;
 877                     $ret = '';
 878                     $iEncStart = false;
 879                 }
 880                 break;
 881             case ' ':
 882                 if ($iEncStart !== false) {
 883                     $cur_l++;
 884                     if ($cur_l > $max_l) {
 885                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 886                         $aRet[] = "=?$default_charset?Q?$ret?=";
 887                         $iOffset = $i;
 888                         $cur_l = 0;
 889                         $ret = '';
 890                         $iEncStart = false;
 891                     } else {
 892                         $ret .= '_';
 893                     }
 894                 }
 895                 break;
 896             default:
 897                 $k = ord($string{$i});
 898                 if ($k > 126) {
 899                     if ($iEncStart === false) {
 900                         // do not start encoding in the middle of a string, also take the rest of the word.
 901                         $sLeadString = substr($string,0,$i);
 902                         $aLeadString = explode(' ',$sLeadString);
 903                         $sToBeEncoded = array_pop($aLeadString);
 904                         $iEncStart = $i - strlen($sToBeEncoded);
 905                         $ret .= $sToBeEncoded;
 906                         $cur_l += strlen($sToBeEncoded);
 907                     }
 908                     $cur_l += 3;
 909                     /* first we add the encoded string that reached it's max size */
 910                     if ($cur_l > ($max_l-2)) {
 911                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 912                         $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
 913                         $cur_l = 3;
 914                         $ret = '';
 915                         $iOffset = $i;
 916                         $iEncStart = $i;
 917                     }
 918                     $enc_init = true;
 919                     $ret .= sprintf("=%02X", $k);
 920                 } else {
 921                     if ($iEncStart !== false) {
 922                         $cur_l++;
 923                         if ($cur_l > $max_l) {
 924                             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 925                             $aRet[] = "=?$default_charset?Q?$ret?=";
 926                             $iEncStart = false;
 927                             $iOffset = $i;
 928                             $cur_l = 0;
 929                             $ret = '';
 930                         } else {
 931                             $ret .= $string{$i};
 932                         }
 933                     }
 934                 }
 935                 break;
 936         }
 937     }
 938
 939     if ($enc_init) {
 940         if ($iEncStart !== false) {
 941             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 942             $aRet[] = "=?$default_charset?Q?$ret?=";
 943         } else {
 944             $aRet[] = substr($string,$iOffset);
 945         }
 946         $string = implode('',$aRet);
 947     }
 948     return $string;
 949 }
 950
 951 /**
 952  * Encodes string according to rfc2047 B encoding header formating rules
 953  *
 954  * It is recommended way to encode headers with character sets that store
 955  * symbols in more than one byte.
 956  *
 957  * Function requires mbstring support. If required mbstring functions are missing,
 958  * function returns false and sets E_USER_WARNING level error message.
 959  *
 960  * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
 961  * that mbstring functions will generate E_WARNING errors, if unsupported
 962  * character set is used. mb_encode_mimeheader function provided by php
 963  * mbstring extension is not used in order to get better control of header
 964  * encoding.
 965  *
 966  * Used php code functions - function_exists(), trigger_error(), strlen()
 967  * (is used with charset names and base64 strings). Used php mbstring
 968  * functions - mb_strlen and mb_substr.
 969  *
 970  * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
 971  * encoding), rfc 2822 (header folding)
 972  *
 973  * @param string $string header string that must be encoded
 974  * @param string $charset character set. Must be supported by mbstring extension.
 975  * Use sq_mb_list_encodings() to detect supported charsets.
 976  * @return string string encoded according to rfc2047 B encoding formating rules
 977  * @since 1.5.1
 978  * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
 979  * @todo Do we want to control max length of header?
 980  * @todo Do we want to control EOL (end-of-line) marker?
 981  * @todo Do we want to translate error message?
 982  */
 983 function encodeHeaderBase64($string,$charset) {
 984     /**
 985      * Check mbstring function requirements.
 986      */
 987     if (! function_exists('mb_strlen') ||
 988         ! function_exists('mb_substr')) {
 989         // set E_USER_WARNING
 990         trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
 991         // return false
 992         return false;
 993     }
 994
 995     // initial return array
 996     $aRet = array();
 997
 998     /**
 999      * header length = 75 symbols max (same as in encodeHeader)
1000      * remove $charset length
1001      * remove =? ? ?= (5 chars)
1002      * remove 2 more chars (\r\n ?)
1003      */
1004     $iMaxLength = 75 - strlen($charset) - 7;
1005
1006     // set first character position
1007     $iStartCharNum = 0;
1008
1009     // loop through all characters. count characters and not bytes.
1010     for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1011         // encode string from starting character to current character.
1012         $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1013
1014         // Check encoded string length
1015         if(strlen($encoded_string)>$iMaxLength) {
1016             // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1017             $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1018
1019             // set new starting character
1020             $iStartCharNum = $iCharNum-1;
1021
1022             // encode last char (in case it is last character in string)
1023             $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1024         } // if string is shorter than max length - add next character
1025     }
1026
1027     // add last encoded string to array
1028     $aRet[] = $encoded_string;
1029
1030     // set initial return string
1031     $sRet = '';
1032
1033     // loop through encoded strings
1034     foreach($aRet as $string) {
1035         // TODO: Do we want to control EOL (end-of-line) marker
1036         if ($sRet!='') $sRet.= " ";
1037
1038         // add header tags and encoded string to return string
1039         $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1040     }
1041
1042     return $sRet;
1043 }
1044
1045 /* This function trys to locate the entity_id of a specific mime element */
1046 function find_ent_id($id, $message) {
1047     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1048         if ($message->entities[$i]->header->type0 == 'multipart')  {
1049             $ret = find_ent_id($id, $message->entities[$i]);
1050         } else {
1051             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
1052 //                if (sq_check_save_extension($message->entities[$i])) {
1053                 return $message->entities[$i]->entity_id;
1054 //                }
1055             } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1056                 /**
1057                  * This is part of a fix for Outlook Express 6.x generating
1058                  * cid URLs without creating content-id headers
1059                  * @@JA - 20050207
1060                  */
1061                 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1062                     return $message->entities[$i]->entity_id;
1063                 }
1064             }
1065         }
1066     }
1067     return $ret;
1068 }
1069
1070 function sq_check_save_extension($message) {
1071     $filename = $message->getFilename();
1072     $ext = substr($filename, strrpos($filename,'.')+1);
1073     $save_extensions = array('jpg','jpeg','gif','png','bmp');
1074     return in_array($ext, $save_extensions);
1075 }
1076
1077
1078 /**
1079  ** HTMLFILTER ROUTINES
1080  */
1081
1082 /**
1083  * This function checks attribute values for entity-encoded values
1084  * and returns them translated into 8-bit strings so we can run
1085  * checks on them.
1086  *
1087  * @param  $attvalue A string to run entity check against.
1088  * @return           Nothing, modifies a reference value.
1089  */
1090 function sq_defang(&$attvalue){
1091     $me = 'sq_defang';
1092     /**
1093      * Skip this if there aren't ampersands or backslashes.
1094      */
1095     if (strpos($attvalue, '&') === false
1096         && strpos($attvalue, '\\') === false){
1097         return;
1098     }
1099     $m = false;
1100     do {
1101         $m = false;
1102         $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1103         $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1104         $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1105     } while ($m == true);
1106     $attvalue = stripslashes($attvalue);
1107 }
1108
1109 /**
1110  * Kill any tabs, newlines, or carriage returns. Our friends the
1111  * makers of the browser with 95% market value decided that it'd
1112  * be funny to make "java[tab]script" be just as good as "javascript".
1113  *
1114  * @param  attvalue  The attribute value before extraneous spaces removed.
1115  * @return attvalue  Nothing, modifies a reference value.
1116  */
1117 function sq_unspace(&$attvalue){
1118     $me = 'sq_unspace';
1119     if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1120         $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1121                                 Array('',   '',   '',   '',   ''), $attvalue);
1122     }
1123 }
1124
1125 /**
1126  * This function returns the final tag out of the tag name, an array
1127  * of attributes, and the type of the tag. This function is called by
1128  * sq_sanitize internally.
1129  *
1130  * @param  $tagname  the name of the tag.
1131  * @param  $attary   the array of attributes and their values
1132  * @param  $tagtype  The type of the tag (see in comments).
1133  * @return           a string with the final tag representation.
1134  */
1135 function sq_tagprint($tagname, $attary, $tagtype){
1136     $me = 'sq_tagprint';
1137
1138     if ($tagtype == 2){
1139         $fulltag = '</' . $tagname . '>';
1140     } else {
1141         $fulltag = '<' . $tagname;
1142         if (is_array($attary) && sizeof($attary)){
1143             $atts = Array();
1144             while (list($attname, $attvalue) = each($attary)){
1145                 array_push($atts, "$attname=$attvalue");
1146             }
1147             $fulltag .= ' ' . join(" ", $atts);
1148         }
1149         if ($tagtype == 3){
1150             $fulltag .= ' /';
1151         }
1152         $fulltag .= '>';
1153     }
1154     return $fulltag;
1155 }
1156
1157 /**
1158  * A small helper function to use with array_walk. Modifies a by-ref
1159  * value and makes it lowercase.
1160  *
1161  * @param  $val a value passed by-ref.
1162  * @return      void since it modifies a by-ref value.
1163  */
1164 function sq_casenormalize(&$val){
1165     $val = strtolower($val);
1166 }
1167
1168 /**
1169  * This function skips any whitespace from the current position within
1170  * a string and to the next non-whitespace value.
1171  *
1172  * @param  $body   the string
1173  * @param  $offset the offset within the string where we should start
1174  *                 looking for the next non-whitespace character.
1175  * @return         the location within the $body where the next
1176  *                 non-whitespace char is located.
1177  */
1178 function sq_skipspace($body, $offset){
1179     $me = 'sq_skipspace';
1180     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1181     if (sizeof($matches{1})){
1182         $count = strlen($matches{1});
1183         $offset += $count;
1184     }
1185     return $offset;
1186 }
1187
1188 /**
1189  * This function looks for the next character within a string.  It's
1190  * really just a glorified "strpos", except it catches if failures
1191  * nicely.
1192  *
1193  * @param  $body   The string to look for needle in.
1194  * @param  $offset Start looking from this position.
1195  * @param  $needle The character/string to look for.
1196  * @return         location of the next occurance of the needle, or
1197  *                 strlen($body) if needle wasn't found.
1198  */
1199 function sq_findnxstr($body, $offset, $needle){
1200     $me  = 'sq_findnxstr';
1201     $pos = strpos($body, $needle, $offset);
1202     if ($pos === FALSE){
1203         $pos = strlen($body);
1204     }
1205     return $pos;
1206 }
1207
1208 /**
1209  * This function takes a PCRE-style regexp and tries to match it
1210  * within the string.
1211  *
1212  * @param  $body   The string to look for needle in.
1213  * @param  $offset Start looking from here.
1214  * @param  $reg    A PCRE-style regex to match.
1215  * @return         Returns a false if no matches found, or an array
1216  *                 with the following members:
1217  *                 - integer with the location of the match within $body
1218  *                 - string with whatever content between offset and the match
1219  *                 - string with whatever it is we matched
1220  */
1221 function sq_findnxreg($body, $offset, $reg){
1222     $me = 'sq_findnxreg';
1223     $matches = Array();
1224     $retarr = Array();
1225     preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1226     if (!isset($matches{0}) || !$matches{0}){
1227         $retarr = false;
1228     } else {
1229         $retarr{0} = $offset + strlen($matches{1});
1230         $retarr{1} = $matches{1};
1231         $retarr{2} = $matches{2};
1232     }
1233     return $retarr;
1234 }
1235
1236 /**
1237  * This function looks for the next tag.
1238  *
1239  * @param  $body   String where to look for the next tag.
1240  * @param  $offset Start looking from here.
1241  * @return         false if no more tags exist in the body, or
1242  *                 an array with the following members:
1243  *                 - string with the name of the tag
1244  *                 - array with attributes and their values
1245  *                 - integer with tag type (1, 2, or 3)
1246  *                 - integer where the tag starts (starting "<")
1247  *                 - integer where the tag ends (ending ">")
1248  *                 first three members will be false, if the tag is invalid.
1249  */
1250 function sq_getnxtag($body, $offset){
1251     $me = 'sq_getnxtag';
1252     if ($offset > strlen($body)){
1253         return false;
1254     }
1255     $lt = sq_findnxstr($body, $offset, "<");
1256     if ($lt == strlen($body)){
1257         return false;
1258     }
1259     /**
1260      * We are here:
1261      * blah blah <tag attribute="value">
1262      * \---------^
1263      */
1264     $pos = sq_skipspace($body, $lt+1);
1265     if ($pos >= strlen($body)){
1266         return Array(false, false, false, $lt, strlen($body));
1267     }
1268     /**
1269      * There are 3 kinds of tags:
1270      * 1. Opening tag, e.g.:
1271      *    <a href="blah">
1272      * 2. Closing tag, e.g.:
1273      *    </a>
1274      * 3. XHTML-style content-less tag, e.g.:
1275      *    <img src="blah" />
1276      */
1277     $tagtype = false;
1278     switch (substr($body, $pos, 1)){
1279         case '/':
1280             $tagtype = 2;
1281             $pos++;
1282             break;
1283         case '!':
1284             /**
1285              * A comment or an SGML declaration.
1286              */
1287             if (substr($body, $pos+1, 2) == "--"){
1288                 $gt = strpos($body, "-->", $pos);
1289                 if ($gt === false){
1290                     $gt = strlen($body);
1291                 } else {
1292                     $gt += 2;
1293                 }
1294                 return Array(false, false, false, $lt, $gt);
1295             } else {
1296                 $gt = sq_findnxstr($body, $pos, ">");
1297                 return Array(false, false, false, $lt, $gt);
1298             }
1299             break;
1300         default:
1301             /**
1302              * Assume tagtype 1 for now. If it's type 3, we'll switch values
1303              * later.
1304              */
1305             $tagtype = 1;
1306             break;
1307     }
1308
1309     $tag_start = $pos;
1310     $tagname = '';
1311     /**
1312      * Look for next [\W-_], which will indicate the end of the tag name.
1313      */
1314     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1315     if ($regary == false){
1316         return Array(false, false, false, $lt, strlen($body));
1317     }
1318     list($pos, $tagname, $match) = $regary;
1319     $tagname = strtolower($tagname);
1320
1321     /**
1322      * $match can be either of these:
1323      * '>'  indicating the end of the tag entirely.
1324      * '\s' indicating the end of the tag name.
1325      * '/'  indicating that this is type-3 xhtml tag.
1326      *
1327      * Whatever else we find there indicates an invalid tag.
1328      */
1329     switch ($match){
1330         case '/':
1331             /**
1332              * This is an xhtml-style tag with a closing / at the
1333              * end, like so: <img src="blah" />. Check if it's followed
1334              * by the closing bracket. If not, then this tag is invalid
1335              */
1336             if (substr($body, $pos, 2) == "/>"){
1337                 $pos++;
1338                 $tagtype = 3;
1339             } else {
1340                 $gt = sq_findnxstr($body, $pos, ">");
1341                 $retary = Array(false, false, false, $lt, $gt);
1342                 return $retary;
1343             }
1344         case '>':
1345             return Array($tagname, false, $tagtype, $lt, $pos);
1346             break;
1347         default:
1348             /**
1349              * Check if it's whitespace
1350              */
1351             if (!preg_match('/\s/', $match)){
1352                 /**
1353                  * This is an invalid tag! Look for the next closing ">".
1354                  */
1355                 $gt = sq_findnxstr($body, $lt, ">");
1356                 return Array(false, false, false, $lt, $gt);
1357             }
1358             break;
1359     }
1360
1361     /**
1362      * At this point we're here:
1363      * <tagname  attribute='blah'>
1364      * \-------^
1365      *
1366      * At this point we loop in order to find all attributes.
1367      */
1368     $attname = '';
1369     $atttype = false;
1370     $attary = Array();
1371
1372     while ($pos <= strlen($body)){
1373         $pos = sq_skipspace($body, $pos);
1374         if ($pos == strlen($body)){
1375             /**
1376              * Non-closed tag.
1377              */
1378             return Array(false, false, false, $lt, $pos);
1379         }
1380         /**
1381          * See if we arrived at a ">" or "/>", which means that we reached
1382          * the end of the tag.
1383          */
1384         $matches = Array();
1385         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1386             /**
1387              * Yep. So we did.
1388              */
1389             $pos += strlen($matches{1});
1390             if ($matches{2} == "/>"){
1391                 $tagtype = 3;
1392                 $pos++;
1393             }
1394             return Array($tagname, $attary, $tagtype, $lt, $pos);
1395         }
1396
1397         /**
1398          * There are several types of attributes, with optional
1399          * [:space:] between members.
1400          * Type 1:
1401          *   attrname[:space:]=[:space:]'CDATA'
1402          * Type 2:
1403          *   attrname[:space:]=[:space:]"CDATA"
1404          * Type 3:
1405          *   attr[:space:]=[:space:]CDATA
1406          * Type 4:
1407          *   attrname
1408          *
1409          * We leave types 1 and 2 the same, type 3 we check for
1410          * '"' and convert to "&quot" if needed, then wrap in
1411          * double quotes. Type 4 we convert into:
1412          * attrname="yes".
1413          */
1414         $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1415         if ($regary == false){
1416             /**
1417              * Looks like body ended before the end of tag.
1418              */
1419             return Array(false, false, false, $lt, strlen($body));
1420         }
1421         list($pos, $attname, $match) = $regary;
1422         $attname = strtolower($attname);
1423         /**
1424          * We arrived at the end of attribute name. Several things possible
1425          * here:
1426          * '>'  means the end of the tag and this is attribute type 4
1427          * '/'  if followed by '>' means the same thing as above
1428          * '\s' means a lot of things -- look what it's followed by.
1429          *      anything else means the attribute is invalid.
1430          */
1431         switch($match){
1432             case '/':
1433                 /**
1434                  * This is an xhtml-style tag with a closing / at the
1435                  * end, like so: <img src="blah" />. Check if it's followed
1436                  * by the closing bracket. If not, then this tag is invalid
1437                  */
1438                 if (substr($body, $pos, 2) == "/>"){
1439                     $pos++;
1440                     $tagtype = 3;
1441                 } else {
1442                     $gt = sq_findnxstr($body, $pos, ">");
1443                     $retary = Array(false, false, false, $lt, $gt);
1444                     return $retary;
1445                 }
1446             case '>':
1447                 $attary{$attname} = '"yes"';
1448                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1449                 break;
1450             default:
1451                 /**
1452                  * Skip whitespace and see what we arrive at.
1453                  */
1454                 $pos = sq_skipspace($body, $pos);
1455                 $char = substr($body, $pos, 1);
1456                 /**
1457                  * Two things are valid here:
1458                  * '=' means this is attribute type 1 2 or 3.
1459                  * \w means this was attribute type 4.
1460                  * anything else we ignore and re-loop. End of tag and
1461                  * invalid stuff will be caught by our checks at the beginning
1462                  * of the loop.
1463                  */
1464                 if ($char == "="){
1465                     $pos++;
1466                     $pos = sq_skipspace($body, $pos);
1467                     /**
1468                      * Here are 3 possibilities:
1469                      * "'"  attribute type 1
1470                      * '"'  attribute type 2
1471                      * everything else is the content of tag type 3
1472                      */
1473                     $quot = substr($body, $pos, 1);
1474                     if ($quot == "'"){
1475                         $regary = sq_findnxreg($body, $pos+1, "\'");
1476                         if ($regary == false){
1477                             return Array(false, false, false, $lt, strlen($body));
1478                         }
1479                         list($pos, $attval, $match) = $regary;
1480                         $pos++;
1481                         $attary{$attname} = "'" . $attval . "'";
1482                     } else if ($quot == '"'){
1483                         $regary = sq_findnxreg($body, $pos+1, '\"');
1484                         if ($regary == false){
1485                             return Array(false, false, false, $lt, strlen($body));
1486                         }
1487                         list($pos, $attval, $match) = $regary;
1488                         $pos++;
1489                         $attary{$attname} = '"' . $attval . '"';
1490                     } else {
1491                         /**
1492                          * These are hateful. Look for \s, or >.
1493                          */
1494                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1495                         if ($regary == false){
1496                             return Array(false, false, false, $lt, strlen($body));
1497                         }
1498                         list($pos, $attval, $match) = $regary;
1499                         /**
1500                          * If it's ">" it will be caught at the top.
1501                          */
1502                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1503                         $attary{$attname} = '"' . $attval . '"';
1504                     }
1505                 } else if (preg_match("|[\w/>]|", $char)) {
1506                     /**
1507                      * That was attribute type 4.
1508                      */
1509                     $attary{$attname} = '"yes"';
1510                 } else {
1511                     /**
1512                      * An illegal character. Find next '>' and return.
1513                      */
1514                     $gt = sq_findnxstr($body, $pos, ">");
1515                     return Array(false, false, false, $lt, $gt);
1516                 }
1517                 break;
1518         }
1519     }
1520     /**
1521      * The fact that we got here indicates that the tag end was never
1522      * found. Return invalid tag indication so it gets stripped.
1523      */
1524     return Array(false, false, false, $lt, strlen($body));
1525 }
1526
1527 /**
1528  * Translates entities into literal values so they can be checked.
1529  *
1530  * @param $attvalue the by-ref value to check.
1531  * @param $regex    the regular expression to check against.
1532  * @param $hex      whether the entites are hexadecimal.
1533  * @return          True or False depending on whether there were matches.
1534  */
1535 function sq_deent(&$attvalue, $regex, $hex=false){
1536     $me = 'sq_deent';
1537     $ret_match = false;
1538     preg_match_all($regex, $attvalue, $matches);
1539     if (is_array($matches) && sizeof($matches[0]) > 0){
1540         $repl = Array();
1541         for ($i = 0; $i < sizeof($matches[0]); $i++){
1542             $numval = $matches[1][$i];
1543             if ($hex){
1544                 $numval = hexdec($numval);
1545             }
1546             $repl{$matches[0][$i]} = chr($numval);
1547         }
1548         $attvalue = strtr($attvalue, $repl);
1549         return true;
1550     } else {
1551         return false;
1552     }
1553 }
1554
1555 /**
1556  * This function runs various checks against the attributes.
1557  *
1558  * @param  $tagname         String with the name of the tag.
1559  * @param  $attary          Array with all tag attributes.
1560  * @param  $rm_attnames     See description for sq_sanitize
1561  * @param  $bad_attvals     See description for sq_sanitize
1562  * @param  $add_attr_to_tag See description for sq_sanitize
1563  * @param  $message         message object
1564  * @param  $id              message id
1565  * @return                  Array with modified attributes.
1566  */
1567 function sq_fixatts($tagname,
1568                     $attary,
1569                     $rm_attnames,
1570                     $bad_attvals,
1571                     $add_attr_to_tag,
1572                     $message,
1573                     $id,
1574                     $mailbox
1575                     ){
1576     $me = 'sq_fixatts';
1577     while (list($attname, $attvalue) = each($attary)){
1578         /**
1579          * See if this attribute should be removed.
1580          */
1581         foreach ($rm_attnames as $matchtag=>$matchattrs){
1582             if (preg_match($matchtag, $tagname)){
1583                 foreach ($matchattrs as $matchattr){
1584                     if (preg_match($matchattr, $attname)){
1585                         unset($attary{$attname});
1586                         continue;
1587                     }
1588                 }
1589             }
1590         }
1591         /**
1592          * Remove any backslashes, entities, and extraneous whitespace.
1593          */
1594         sq_defang($attvalue);
1595         sq_unspace($attvalue);
1596
1597         /**
1598          * Now let's run checks on the attvalues.
1599          * I don't expect anyone to comprehend this. If you do,
1600          * get in touch with me so I can drive to where you live and
1601          * shake your hand personally. :)
1602          */
1603         foreach ($bad_attvals as $matchtag=>$matchattrs){
1604             if (preg_match($matchtag, $tagname)){
1605                 foreach ($matchattrs as $matchattr=>$valary){
1606                     if (preg_match($matchattr, $attname)){
1607                         /**
1608                          * There are two arrays in valary.
1609                          * First is matches.
1610                          * Second one is replacements
1611                          */
1612                         list($valmatch, $valrepl) = $valary;
1613                         $newvalue =
1614                             preg_replace($valmatch, $valrepl, $attvalue);
1615                         if ($newvalue != $attvalue){
1616                             $attary{$attname} = $newvalue;
1617                         }
1618                     }
1619                 }
1620             }
1621         }
1622
1623         /**
1624          * Replace empty src tags with the blank image.  src is only used
1625          * for frames, images, and image inputs.  Doing a replace should
1626          * not affect them working as should be, however it will stop
1627          * IE from being kicked off when src for img tags are not set
1628          */
1629         if (($attname == 'src') && ($attvalue == '""')) {
1630             $attary{$attname} = '"' . SM_PATH . 'images/blank.png"';
1631         }
1632
1633         /**
1634          * Turn cid: urls into http-friendly ones.
1635          */
1636         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1637             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1638         }
1639
1640         /**
1641          * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1642          * One day MS might actually make it match something useful, for now, falling
1643          * back to using cid2http, so we can grab the blank.png.
1644          */
1645         if (preg_match("/^[\'\"]\s*outbind:\/\//si", $attvalue)) {
1646             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1647         }
1648
1649     }
1650     /**
1651      * See if we need to append any attributes to this tag.
1652      */
1653     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1654         if (preg_match($matchtag, $tagname)){
1655             $attary = array_merge($attary, $addattary);
1656         }
1657     }
1658     return $attary;
1659 }
1660
1661 /**
1662  * This function edits the style definition to make them friendly and
1663  * usable in SquirrelMail.
1664  *
1665  * @param  $message  the message object
1666  * @param  $id       the message id
1667  * @param  $content  a string with whatever is between <style> and </style>
1668  * @param  $mailbox  the message mailbox
1669  * @return           a string with edited content.
1670  */
1671 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1672     global $view_unsafe_images;
1673     $me = 'sq_fixstyle';
1674     $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1675     if ($ret == FALSE){
1676         return array(FALSE, strlen($body));
1677     }
1678     $newpos = $ret[0] + strlen($ret[2]);
1679     $content = $ret[1];
1680     /**
1681     * First look for general BODY style declaration, which would be
1682     * like so:
1683     * body {background: blah-blah}
1684     * and change it to .bodyclass so we can just assign it to a <div>
1685     */
1686     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1687     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1688     /**
1689     * Fix url('blah') declarations.
1690     */
1691     //   $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1692     //                           "url(\\1$secremoveimg\\2)", $content);
1693     // remove NUL
1694     $content = str_replace("\0", "", $content);
1695     // NB I insert NUL characters to keep to avoid an infinite loop. They are removed after the loop.
1696     while (preg_match("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si", $content, $matches)) {
1697         $sProto = strtolower($matches[1]);
1698         switch ($sProto) {
1699           /**
1700            * Fix url('https*://.*) declarations but only if $view_unsafe_images
1701            * is false.
1702            */
1703           case 'https':
1704           case 'http':
1705             if (!$view_unsafe_images){
1706                 $sExpr = "/url\s*\(\s*([\'\"])\s*$sProto*:.*?([\'\"])\s*\)/si";
1707                 $content = preg_replace($sExpr, "u\0r\0l(\\1$secremoveimg\\2)", $content);
1708             }
1709             break;
1710           /**
1711            * Fix urls that refer to cid:
1712            */
1713           case 'cid':
1714             $cidurl = 'cid:'. $matches[2];
1715             $httpurl = sq_cid2http($message, $id, $cidurl, $mailbox);
1716             $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1717                                 "u\0r\0l($httpurl)", $content);
1718             break;
1719           default:
1720             /**
1721              * replace url with protocol other then the white list
1722              * http,https and cid by an empty string.
1723              */
1724             $content = preg_replace("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si",
1725                                 "", $content);
1726             break;
1727         }
1728         break;
1729     }
1730     // remove NUL
1731     $content = str_replace("\0", "", $content);
1732
1733    /**
1734     * Remove any backslashes, entities, and extraneous whitespace.
1735     */
1736     $contentTemp = $content;
1737     sq_defang($contentTemp);
1738     sq_unspace($contentTemp);
1739
1740     /**
1741      * Fix stupid css declarations which lead to vulnerabilities
1742      * in IE.
1743      */
1744     $match   = Array('/expression/i',
1745                     '/behaviou*r/i',
1746                     '/binding/i',
1747                     '/include-source/i');
1748     $replace = Array('idiocy', 'idiocy', 'idiocy', 'idiocy');
1749     $contentNew = preg_replace($match, $replace, $contentTemp);
1750     if ($contentNew !== $contentTemp) {
1751         // insecure css declarations are used. From now on we don't care
1752         // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
1753         $content = $contentNew;
1754     }
1755     return array($content, $newpos);
1756 }
1757
1758
1759 /**
1760  * This function converts cid: url's into the ones that can be viewed in
1761  * the browser.
1762  *
1763  * @param  $message  the message object
1764  * @param  $id       the message id
1765  * @param  $cidurl   the cid: url.
1766  * @param  $mailbox  the message mailbox
1767  * @return           a string with a http-friendly url
1768  */
1769 function sq_cid2http($message, $id, $cidurl, $mailbox){
1770     /**
1771      * Get rid of quotes.
1772      */
1773     $quotchar = substr($cidurl, 0, 1);
1774     if ($quotchar == '"' || $quotchar == "'"){
1775         $cidurl = str_replace($quotchar, "", $cidurl);
1776     } else {
1777         $quotchar = '';
1778     }
1779     $cidurl = substr(trim($cidurl), 4);
1780
1781     $match_str = '/\{.*?\}\//';
1782     $str_rep = '';
1783     $cidurl = preg_replace($match_str, $str_rep, $cidurl);
1784
1785     $linkurl = find_ent_id($cidurl, $message);
1786     /* in case of non-save cid links $httpurl should be replaced by a sort of
1787        unsave link image */
1788     $httpurl = '';
1789
1790     /**
1791      * This is part of a fix for Outlook Express 6.x generating
1792      * cid URLs without creating content-id headers. These images are
1793      * not part of the multipart/related html mail. The html contains
1794      * <img src="cid:{some_id}/image_filename.ext"> references to
1795      * attached images with as goal to render them inline although
1796      * the attachment disposition property is not inline.
1797      */
1798
1799     if (empty($linkurl)) {
1800         if (preg_match('/{.*}\//', $cidurl)) {
1801             $cidurl = preg_replace('/{.*}\//','', $cidurl);
1802             if (!empty($cidurl)) {
1803                 $linkurl = find_ent_id($cidurl, $message);
1804             }
1805         }
1806     }
1807
1808     if (!empty($linkurl)) {
1809         $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1810             "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1811             '&amp;ent_id=' . $linkurl . $quotchar;
1812     } else {
1813         /**
1814          * If we couldn't generate a proper img url, drop in a blank image
1815          * instead of sending back empty, otherwise it causes unusual behaviour
1816          */
1817         $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
1818     }
1819
1820     return $httpurl;
1821 }
1822
1823 /**
1824  * This function changes the <body> tag into a <div> tag since we
1825  * can't really have a body-within-body.
1826  *
1827  * @param  $attary   an array of attributes and values of <body>
1828  * @param  $mailbox  mailbox we're currently reading (for cid2http)
1829  * @param  $message  current message (for cid2http)
1830  * @param  $id       current message id (for cid2http)
1831  * @return           a modified array of attributes to be set for <div>
1832  */
1833 function sq_body2div($attary, $mailbox, $message, $id){
1834     $me = 'sq_body2div';
1835     $divattary = Array('class' => "'bodyclass'");
1836     $text = '#000000';
1837     $has_bgc_stl = $has_txt_stl = false;
1838     $styledef = '';
1839     if (is_array($attary) && sizeof($attary) > 0){
1840         foreach ($attary as $attname=>$attvalue){
1841             $quotchar = substr($attvalue, 0, 1);
1842             $attvalue = str_replace($quotchar, "", $attvalue);
1843             switch ($attname){
1844                 case 'background':
1845                     $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1846                     $styledef .= "background-image: url('$attvalue'); ";
1847                     break;
1848                 case 'bgcolor':
1849                     $has_bgc_stl = true;
1850                     $styledef .= "background-color: $attvalue; ";
1851                     break;
1852                 case 'text':
1853                     $has_txt_stl = true;
1854                     $styledef .= "color: $attvalue; ";
1855                     break;
1856             }
1857         }
1858         // Outlook defines a white bgcolor and no text color. This can lead to
1859         // white text on a white bg with certain themes.
1860         if ($has_bgc_stl && !$has_txt_stl) {
1861             $styledef .= "color: $text; ";
1862         }
1863         if (strlen($styledef) > 0){
1864             $divattary{"style"} = "\"$styledef\"";
1865         }
1866     }
1867     return $divattary;
1868 }
1869
1870 /**
1871  * This is the main function and the one you should actually be calling.
1872  * There are several variables you should be aware of an which need
1873  * special description.
1874  *
1875  * Since the description is quite lengthy, see it here:
1876  * http://linux.duke.edu/projects/mini/htmlfilter/
1877  *
1878  * @param $body                 the string with HTML you wish to filter
1879  * @param $tag_list             see description above
1880  * @param $rm_tags_with_content see description above
1881  * @param $self_closing_tags    see description above
1882  * @param $force_tag_closing    see description above
1883  * @param $rm_attnames          see description above
1884  * @param $bad_attvals          see description above
1885  * @param $add_attr_to_tag      see description above
1886  * @param $message              message object
1887  * @param $id                   message id
1888  * @return                      sanitized html safe to show on your pages.
1889  */
1890 function sq_sanitize($body,
1891                      $tag_list,
1892                      $rm_tags_with_content,
1893                      $self_closing_tags,
1894                      $force_tag_closing,
1895                      $rm_attnames,
1896                      $bad_attvals,
1897                      $add_attr_to_tag,
1898                      $message,
1899                      $id,
1900                      $mailbox
1901                      ){
1902     $me = 'sq_sanitize';
1903     $rm_tags = array_shift($tag_list);
1904     /**
1905      * Normalize rm_tags and rm_tags_with_content.
1906      */
1907     @array_walk($tag_list, 'sq_casenormalize');
1908     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1909     @array_walk($self_closing_tags, 'sq_casenormalize');
1910     /**
1911      * See if tag_list is of tags to remove or tags to allow.
1912      * false  means remove these tags
1913      * true   means allow these tags
1914      */
1915     $curpos = 0;
1916     $open_tags = Array();
1917     $trusted = "\n<!-- begin sanitized html -->\n";
1918     $skip_content = false;
1919     /**
1920      * Take care of netscape's stupid javascript entities like
1921      * &{alert('boo')};
1922      */
1923     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1924
1925     while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1926         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1927         $free_content = substr($body, $curpos, $lt-$curpos);
1928         /**
1929          * Take care of <style>
1930          */
1931         if ($tagname == "style" && $tagtype == 1){
1932             list($free_content, $curpos) =
1933                 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
1934             if ($free_content != FALSE){
1935                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1936                 $trusted .= $free_content;
1937                 $trusted .= sq_tagprint($tagname, false, 2);
1938             }
1939             continue;
1940         }
1941         if ($skip_content == false){
1942             $trusted .= $free_content;
1943         }
1944         if ($tagname != FALSE){
1945             if ($tagtype == 2){
1946                 if ($skip_content == $tagname){
1947                     /**
1948                      * Got to the end of tag we needed to remove.
1949                      */
1950                     $tagname = false;
1951                     $skip_content = false;
1952                 } else {
1953                     if ($skip_content == false){
1954                         if ($tagname == "body"){
1955                             $tagname = "div";
1956                         }
1957                         if (isset($open_tags{$tagname}) &&
1958                                 $open_tags{$tagname} > 0){
1959                             $open_tags{$tagname}--;
1960                         } else {
1961                             $tagname = false;
1962                         }
1963                     }
1964                 }
1965             } else {
1966                 /**
1967                  * $rm_tags_with_content
1968                  */
1969                 if ($skip_content == false){
1970                     /**
1971                      * See if this is a self-closing type and change
1972                      * tagtype appropriately.
1973                      */
1974                     if ($tagtype == 1
1975                             && in_array($tagname, $self_closing_tags)){
1976                         $tagtype = 3;
1977                     }
1978                     /**
1979                      * See if we should skip this tag and any content
1980                      * inside it.
1981                      */
1982                     if ($tagtype == 1 &&
1983                             in_array($tagname, $rm_tags_with_content)){
1984                         $skip_content = $tagname;
1985                     } else {
1986                         if (($rm_tags == false
1987                                     && in_array($tagname, $tag_list)) ||
1988                                 ($rm_tags == true &&
1989                                  !in_array($tagname, $tag_list))){
1990                             $tagname = false;
1991                         } else {
1992                             /**
1993                              * Convert body into div.
1994                              */
1995                             if ($tagname == "body"){
1996                                 $tagname = "div";
1997                                 $attary = sq_body2div($attary, $mailbox,
1998                                         $message, $id);
1999                             }
2000                             if ($tagtype == 1){
2001                                 if (isset($open_tags{$tagname})){
2002                                     $open_tags{$tagname}++;
2003                                 } else {
2004                                     $open_tags{$tagname}=1;
2005                                 }
2006                             }
2007                             /**
2008                              * This is where we run other checks.
2009                              */
2010                             if (is_array($attary) && sizeof($attary) > 0){
2011                                 $attary = sq_fixatts($tagname,
2012                                                      $attary,
2013                                                      $rm_attnames,
2014                                                      $bad_attvals,
2015                                                      $add_attr_to_tag,
2016                                                      $message,
2017                                                      $id,
2018                                                      $mailbox
2019                                                      );
2020                             }
2021                         }
2022                     }
2023                 }
2024             }
2025             if ($tagname != false && $skip_content == false){
2026                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2027             }
2028         }
2029         $curpos = $gt+1;
2030     }
2031     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2032     if ($force_tag_closing == true){
2033         foreach ($open_tags as $tagname=>$opentimes){
2034             while ($opentimes > 0){
2035                 $trusted .= '</' . $tagname . '>';
2036                 $opentimes--;
2037             }
2038         }
2039         $trusted .= "\n";
2040     }
2041     $trusted .= "<!-- end sanitized html -->\n";
2042     return $trusted;
2043 }
2044
2045 /**
2046  * This is a wrapper function to call html sanitizing routines.
2047  *
2048  * @param  $body  the body of the message
2049  * @param  $id    the id of the message
2050
2051  * @param  $message
2052  * @param  $mailbox
2053  * @param  boolean $take_mailto_links When TRUE, converts mailto: links
2054  *                                    into internal SM compose links
2055  *                                    (optional; default = TRUE)
2056  * @return        a string with html safe to display in the browser.
2057  */
2058 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2059
2060     require_once(SM_PATH . 'functions/url_parser.php');  // for $MailTo_PReg_Match
2061
2062     global $attachment_common_show_images, $view_unsafe_images,
2063            $has_unsafe_images;
2064     /**
2065      * Don't display attached images in HTML mode.
2066      */
2067     $attachment_common_show_images = false;
2068     $tag_list = Array(
2069             false,
2070             "object",
2071             "meta",
2072             "html",
2073             "head",
2074             "base",
2075             "link",
2076             "frame",
2077             "iframe",
2078             "plaintext",
2079             "marquee"
2080             );
2081
2082     $rm_tags_with_content = Array(
2083             "script",
2084             "applet",
2085             "embed",
2086             "title",
2087             "frameset",
2088             "xmp",
2089             "xml"
2090             );
2091
2092     $self_closing_tags =  Array(
2093             "img",
2094             "br",
2095             "hr",
2096             "input",
2097             "outbind"
2098             );
2099
2100     $force_tag_closing = true;
2101
2102     $rm_attnames = Array(
2103             "/.*/" =>
2104             Array(
2105                 "/target/i",
2106                 "/^on.*/i",
2107                 "/^dynsrc/i",
2108                 "/^data.*/i",
2109                 "/^lowsrc.*/i"
2110                 )
2111             );
2112
2113     $secremoveimg = "../images/" . _("sec_remove_eng.png");
2114     $bad_attvals = Array(
2115             "/.*/" =>
2116             Array(
2117                 "/^src|background/i" =>
2118                 Array(
2119                     Array(
2120                         "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2121                         "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2122                         "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2123                         ),
2124                     Array(
2125                         "\\1$secremoveimg\\2",
2126                         "\\1$secremoveimg\\2",
2127                         "\\1$secremoveimg\\2",
2128                         )
2129                     ),
2130                 "/^href|action/i" =>
2131                 Array(
2132                     Array(
2133                         "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2134                         "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2135                         "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2136                         ),
2137                     Array(
2138                         "\\1#\\1",
2139                         "\\1#\\1",
2140                         "\\1#\\1"
2141                         )
2142                     ),
2143         "/^style/i" =>
2144             Array(
2145                 Array(
2146                     "/expression/i",
2147                     "/binding/i",
2148                     "/behaviou*r/i",
2149                     "/include-source/i",
2150                     "/position\s*:\s*absolute/i",
2151                     "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2152                     "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2153                     "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2154                     "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2155                     ),
2156                 Array(
2157                     "idiocy",
2158                     "idiocy",
2159                     "idiocy",
2160                     "idiocy",
2161                     "",
2162                     "url(\\1#\\1)",
2163                     "url(\\1#\\1)",
2164                     "url(\\1#\\1)",
2165                     "\\1:url(\\2#\\3)"
2166                     )
2167                 )
2168             )
2169         );
2170     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
2171         $view_unsafe_images = false;
2172     }
2173     if (!$view_unsafe_images){
2174         /**
2175          * Remove any references to http/https if view_unsafe_images set
2176          * to false.
2177          */
2178         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2179                 '/^([\'\"])\s*https*:.*([\'\"])/si');
2180         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2181                 "\\1$secremoveimg\\1");
2182         array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2183                 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2184         array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2185                 "url(\\1$secremoveimg\\1)");
2186     }
2187
2188     $add_attr_to_tag = Array(
2189             "/^a$/i" =>
2190             Array('target'=>'"_blank"',
2191                 'title'=>'"'._("This external link will open in a new window").'"'
2192                 )
2193             );
2194     $trusted = sq_sanitize($body,
2195                            $tag_list,
2196                            $rm_tags_with_content,
2197                            $self_closing_tags,
2198                            $force_tag_closing,
2199                            $rm_attnames,
2200                            $bad_attvals,
2201                            $add_attr_to_tag,
2202                            $message,
2203                            $id,
2204                            $mailbox
2205                            );
2206     if (preg_match("|$secremoveimg|i", $trusted)){
2207         $has_unsafe_images = true;
2208     }
2209
2210     // we want to parse mailto's in HTML output, change to SM compose links
2211     // this is a modified version of code from url_parser.php... but Marc is
2212     // right: we need a better filtering implementation; adding this randomly
2213     // here is not a great solution
2214     //
2215     if ($take_mailto_links) {
2216         // parseUrl($trusted);   // this even parses URLs inside of tags... too aggressive
2217         global $MailTo_PReg_Match;
2218         $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1);
2219         if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2220             foreach ($regs[0] as $i => $mailto_before) {
2221                 $mailto_params = $regs[10][$i];
2222                 // get rid of any tailing quote since we have to add send_to to the end
2223                 //
2224                 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2225                     $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2226                 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2227                     $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2228
2229                 if ($regs[1][$i]) {    //if there is an email addr before '?', we need to merge it with the params
2230                     $to = 'to=' . $regs[1][$i];
2231                     if (strpos($mailto_params, 'to=') > -1)    //already a 'to='
2232                         $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2233                     else {
2234                         if ($mailto_params)    //already some params, append to them
2235                             $mailto_params .= '&amp;' . $to;
2236                         else
2237                             $mailto_params .= '?' . $to;
2238                     }
2239                 }
2240
2241                 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2242
2243                 // we'll already have target=_blank, no need to allow comp_in_new
2244                 // here (which would be a lot more work anyway)
2245                 //
2246                 global $compose_new_win;
2247                 $temp_comp_in_new = $compose_new_win;
2248                 $compose_new_win = 0;
2249                 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2250                 $compose_new_win = $temp_comp_in_new;
2251
2252                 // remove <a href=" and anything after the next quote (we only
2253                 // need the uri, not the link HTML) in compose uri
2254                 //
2255                 $comp_uri = substr($comp_uri, 9);
2256                 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2257                 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2258             }
2259         }
2260     }
2261
2262     return $trusted;
2263 }
2264
2265 /**
2266  * function SendDownloadHeaders - send file to the browser
2267  *
2268  * Original Source: SM core src/download.php
2269  * moved here to make it available to other code, and separate
2270  * front end from back end functionality.
2271  *
2272  * @param string $type0 first half of mime type
2273  * @param string $type1 second half of mime type
2274  * @param string $filename filename to tell the browser for downloaded file
2275  * @param boolean $force whether to force the download dialog to pop
2276  * @param optional integer $filesize send the Content-Header and length to the browser
2277  * @return void
2278  */
2279 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2280     global $languages, $squirrelmail_language;
2281     $isIE = $isIE6 = 0;
2282
2283     sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2284
2285     if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2286             strstr($HTTP_USER_AGENT, 'Opera') === false) {
2287         $isIE = 1;
2288     }
2289
2290     if (strstr($HTTP_USER_AGENT, 'compatible; MSIE 6') !== false &&
2291             strstr($HTTP_USER_AGENT, 'Opera') === false) {
2292         $isIE6 = 1;
2293     }
2294
2295     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2296             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2297         $filename =
2298             call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2299     } else {
2300         $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
2301     }
2302
2303     // A Pox on Microsoft and it's Internet Explorer!
2304     //
2305     // IE has lots of bugs with file downloads.
2306     // It also has problems with SSL.  Both of these cause problems
2307     // for us in this function.
2308     //
2309     // See this article on Cache Control headers and SSL
2310     // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2311     //
2312     // The best thing you can do for IE is to upgrade to the latest
2313     // version
2314     //set all the Cache Control Headers for IE
2315     if ($isIE) {
2316         $filename=rawurlencode($filename);
2317         header ("Pragma: public");
2318         header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2319         header ("Cache-Control: post-check=0, pre-check=0", false);
2320         header ("Cache-Control: private");
2321
2322         //set the inline header for IE, we'll add the attachment header later if we need it
2323         header ("Content-Disposition: inline; filename=$filename");
2324     }
2325
2326     if (!$force) {
2327         // Try to show in browser window
2328         header ("Content-Disposition: inline; filename=\"$filename\"");
2329         header ("Content-Type: $type0/$type1; name=\"$filename\"");
2330     } else {
2331         // Try to pop up the "save as" box
2332
2333         // IE makes this hard.  It pops up 2 save boxes, or none.
2334         // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2335         // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2336         // But, according to Microsoft, it is "RFC compliant but doesn't
2337         // take into account some deviations that allowed within the
2338         // specification."  Doesn't that mean RFC non-compliant?
2339         // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2340
2341         // all browsers need the application/octet-stream header for this
2342         header ("Content-Type: application/octet-stream; name=\"$filename\"");
2343
2344         // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2345         // Do not have quotes around filename, but that applied to
2346         // "attachment"... does it apply to inline too?
2347         header ("Content-Disposition: attachment; filename=\"$filename\"");
2348
2349         if ($isIE && !$isIE6) {
2350             // This combination seems to work mostly.  IE 5.5 SP 1 has
2351             // known issues (see the Microsoft Knowledge Base)
2352
2353             // This works for most types, but doesn't work with Word files
2354             header ("Content-Type: application/download; name=\"$filename\"");
2355
2356             // These are spares, just in case.  :-)
2357             //header("Content-Type: $type0/$type1; name=\"$filename\"");
2358             //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2359             //header("Content-Type: application/octet-stream; name=\"$filename\"");
2360         } else {
2361             // another application/octet-stream forces download for Netscape
2362             header ("Content-Type: application/octet-stream; name=\"$filename\"");
2363         }
2364     }
2365
2366     //send the content-length header if the calling function provides it
2367     if ($filesize > 0) {
2368         header("Content-Length: $filesize");
2369     }
2370
2371 }  // end fn SendDownloadHeaders
2372
2373 ?>