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; 1999-2006 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,
 338            $use_iframe, $iframe_height, $download_and_unsafe_link;
 339
 340     // workaround for not updated config.php
 341     if (! isset($use_iframe)) $use_iframe = false;
 342
 343     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
 344         $view_unsafe_images = false;
 345     }
 346
 347     $body = '';
 348     $urlmailbox = urlencode($mailbox);
 349     $body_message = getEntity($message, $ent_num);
 350     if (($body_message->header->type0 == 'text') ||
 351             ($body_message->header->type0 == 'rfc822')) {
 352         $body = mime_fetch_body ($imap_stream, $id, $ent_num);
 353         $body = decodeBody($body, $body_message->header->encoding);
 354
 355         if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 356                 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
 357             if (mb_detect_encoding($body) != 'ASCII') {
 358                 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
 359             }
 360         }
 361         $hookResults = do_hook("message_body", $body);
 362         $body = $hookResults[1];
 363
 364         /* If there are other types that shouldn't be formatted, add
 365          * them here.
 366          */
 367
 368         if ($body_message->header->type1 == 'html') {
 369             if ($show_html_default <> 1) {
 370                 $entity_conv = array('&nbsp;' => ' ',
 371                                      '<p>'    => "\n",
 372                                      '<P>'    => "\n",
 373                                      '<br>'   => "\n",
 374                                      '<BR>'   => "\n",
 375                                      '<br />' => "\n",
 376                                      '<BR />' => "\n",
 377                                      '&gt;'   => '>',
 378                                      '&lt;'   => '<');
 379                 $body = strtr($body, $entity_conv);
 380                 $body = strip_tags($body);
 381                 $body = trim($body);
 382                 translateText($body, $wrap_at,
 383                         $body_message->header->getParameter('charset'));
 384             } elseif ($use_iframe && ! $clean) {
 385                 // $clean is used to remove iframe in printable view.
 386
 387                 /**
 388                  * If we don't add html message between iframe tags,
 389                  * we must detect unsafe images and modify $has_unsafe_images.
 390                  */
 391                 $html_body = magicHTML($body, $id, $message, $mailbox);
 392                 // Convert character set in order to display html mails in different character set
 393                 $html_body = charset_decode($body_message->header->getParameter('charset'),$html_body,false,true);
 394
 395                 // creating iframe url
 396                 $iframeurl=sqm_baseuri().'src/view_html.php?'
 397                     . 'mailbox=' . $urlmailbox
 398                     . '&amp;passed_id=' . $id
 399                     . '&amp;ent_id=' . $ent_num
 400                     . '&amp;view_unsafe_images=' . (int) $view_unsafe_images;
 401
 402                 // adding warning message
 403                 $body = html_tag('div',_("Viewing HTML formatted email"),'center');
 404
 405                 /**
 406                  * height can't be set to 100%, because it does not work as expected when
 407                  * iframe is inside the table. Browsers do not create full height objects
 408                  * even when iframe is not nested. Maybe there is some way to get full size
 409                  * with CSS. Tested in firefox 1.02 and opera 7.53
 410                  *
 411                  * width="100%" does not work as expected, when table width is not set (automatic)
 412                  *
 413                  * tokul: I think <iframe> are safer sandbox than <object>. Objects might
 414                  * need special handling for IE and IE6SP2.
 415                  */
 416                 $body.= "<div><iframe name=\"message_frame\" width=\"100%\" height=\"$iframe_height\" src=\"$iframeurl\""
 417                     .' frameborder="1" marginwidth="0" marginheight="0" scrolling="auto">' . "\n";
 418
 419                 // Message for browsers without iframe support
 420                 //$body.= _("Your browser does not support inline frames.
 421                 // You can view HTML formated message by following below link.");
 422                 //$body.= "<br /><a href=\"$iframeurl\">"._("View HTML Message")."</a>";
 423
 424                 // if browser can't render iframe, it renders html message.
 425                 $body.= $html_body;
 426
 427                 // close iframe
 428                 $body.="</iframe></div>\n";
 429             } else {
 430                 // old way of html rendering
 431                 $body = magicHTML($body, $id, $message, $mailbox);
 432                 /**
 433                  * convert character set. charset_decode does not remove html special chars
 434                  * applied by magicHTML functions and does not sanitize them second time if
 435                  * fourth argument is true.
 436                  */
 437                 $body = charset_decode($body_message->header->getParameter('charset'),$body,false,true);
 438             }
 439         } else {
 440             translateText($body, $wrap_at,
 441                     $body_message->header->getParameter('charset'));
 442         }
 443
 444         // if this is the clean display (i.e. printer friendly), stop here.
 445         if ( $clean ) {
 446             return $body;
 447         }
 448
 449         $download_and_unsafe_link = '';
 450
 451         $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
 452             '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
 453             '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 454         if (isset($passed_ent_id)) {
 455             $link .= '&amp;passed_ent_id='.$passed_ent_id;
 456         }
 457         $download_and_unsafe_link .= '&nbsp;|&nbsp;<a href="download.php?absolute_dl=true&amp;' .
 458             $link . '">' . _("Download this as a file") .  '</a>';
 459         if ($view_unsafe_images) {
 460             $text = _("Hide Unsafe Images");
 461         } else {
 462             if (isset($has_unsafe_images) && $has_unsafe_images) {
 463                 $link .= '&amp;view_unsafe_images=1';
 464                 $text = _("View Unsafe Images");
 465             } else {
 466                 $text = '';
 467             }
 468         }
 469         if($text != '') {
 470             $download_and_unsafe_link .= '&nbsp;|&nbsp;<a href="read_body.php?' . $link . '">' . $text . '</a>';
 471         }
 472     }
 473     return $body;
 474 }
 475
 476 /**
 477  * Displays attachment links and information
 478  *
 479  * Since 1.3.0 function is not included in formatBody() call.
 480  *
 481  * Since 1.0.2 uses attachment $type0/$type1 hook.
 482  * Since 1.2.5 uses attachment $type0/* hook.
 483  * Since 1.5.0 uses attachments_bottom hook.
 484  *
 485  * @param object $message SquirrelMail message object
 486  * @param array $exclude_id message parts that are not attachments.
 487  * @param string $mailbox mailbox name
 488  * @param integer $id message id
 489  * @return string html formated attachment information.
 490  */
 491 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 492     global $where, $what, $startMessage, $color, $passed_ent_id;
 493
 494     $att_ar = $message->getAttachments($exclude_id);
 495
 496     if (!count($att_ar)) return '';
 497
 498     $attachments = '';
 499
 500     $urlMailbox = urlencode($mailbox);
 501
 502     foreach ($att_ar as $att) {
 503         $ent = $att->entity_id;
 504         $header = $att->header;
 505         $type0 = strtolower($header->type0);
 506         $type1 = strtolower($header->type1);
 507         $name = '';
 508         $links['download link']['text'] = _("Download");
 509         $links['download link']['href'] = sqm_baseuri() .
 510             "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 511         if ($type0 =='message' && $type1 == 'rfc822') {
 512             $default_page = sqm_baseuri() . 'src/read_body.php';
 513             $rfc822_header = $att->rfc822_header;
 514             $filename = $rfc822_header->subject;
 515             if (trim( $filename ) == '') {
 516                 $filename = 'untitled-[' . $ent . ']' ;
 517             }
 518             $from_o = $rfc822_header->from;
 519             if (is_object($from_o)) {
 520                 $from_name = decodeHeader($from_o->getAddress(false));
 521             } else {
 522                 $from_name = _("Unknown sender");
 523             }
 524             $description = $from_name;
 525         } else {
 526             $default_page = sqm_baseuri() . 'src/download.php';
 527             $filename = $att->getFilename();
 528             if ($header->description) {
 529                 $description = decodeHeader($header->description);
 530             } else {
 531                 $description = '';
 532             }
 533         }
 534
 535         $display_filename = $filename;
 536         if (isset($passed_ent_id)) {
 537             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 538         } else {
 539             $passed_ent_id_link = '';
 540         }
 541         $defaultlink = $default_page . "?startMessage=$startMessage"
 542             . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 543             . '&amp;ent_id='.$ent.$passed_ent_id_link;
 544         if ($where && $what) {
 545             $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 546         }
 547
 548         /* This executes the attachment hook with a specific MIME-type.
 549          * If that doesn't have results, it tries if there's a rule
 550          * for a more generic type.
 551          */
 552         $hookresults = do_hook("attachment $type0/$type1", $links,
 553                 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 554                 $display_filename, $where, $what);
 555         if(count($hookresults[1]) <= 1) {
 556             $hookresults = do_hook("attachment $type0/*", $links,
 557                     $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 558                     $display_filename, $where, $what);
 559         }
 560
 561         $links = $hookresults[1];
 562         $defaultlink = $hookresults[6];
 563
 564         $attachments .= '<tr><td>' .
 565             '<a href="'.$defaultlink.'">'.decodeHeader($display_filename).'</a>&nbsp;</td>' .
 566             '<td><small><b>' . show_readable_size($header->size) .
 567             '</b>&nbsp;&nbsp;</small></td>' .
 568             '<td><small>[ '.htmlspecialchars($type0).'/'.htmlspecialchars($type1).' ]&nbsp;</small></td>' .
 569             '<td><small>';
 570         $attachments .= '<b>' . $description . '</b>';
 571         $attachments .= '</small></td><td><small>&nbsp;';
 572
 573         $skipspaces = 1;
 574         foreach ($links as $val) {
 575             if ($skipspaces) {
 576                 $skipspaces = 0;
 577             } else {
 578                 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
 579             }
 580             $attachments .= '<a href="' . $val['href'] . '">'
 581                 . (isset($val['text']) && !empty($val['text']) ? $val['text'] : '')
 582                 . (isset($val['extra']) && !empty($val['extra']) ? $val['extra'] : '')
 583                 . '</a>';
 584         }
 585         unset($links);
 586         $attachments .= "</td></tr>\n";
 587     }
 588     $attachmentadd = do_hook_function('attachments_bottom',$attachments);
 589     if ($attachmentadd != '')
 590         $attachments = $attachmentadd;
 591     return $attachments;
 592 }
 593
 594 function sqimap_base64_decode(&$string) {
 595
 596     // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
 597     // fly decoding (to reduce memory usage) you have to check if the
 598     // data has incomplete pairs
 599
 600     // Remove the noise in order to check if the 4 bytes pairs are complete
 601     $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
 602
 603     $sStringRem = '';
 604     $iMod = strlen($string) % 4;
 605     if ($iMod) {
 606         $sStringRem = substr($string,-$iMod);
 607         // Check if $sStringRem contains padding characters
 608         if (substr($sStringRem,-1) != '=') {
 609             $string = substr($string,0,-$iMod);
 610         } else {
 611             $sStringRem = '';
 612         }
 613     }
 614     $string = base64_decode($string);
 615     return $sStringRem;
 616 }
 617
 618 /**
 619  * Decodes encoded message body
 620  *
 621  * This function decodes the body depending on the encoding type.
 622  * Currently quoted-printable and base64 encodings are supported.
 623  * decode_body hook was added to this function in 1.4.2/1.5.0
 624  * @param string $body encoded message body
 625  * @param string $encoding used encoding
 626  * @return string decoded string
 627  * @since 1.0
 628  */
 629 function decodeBody($body, $encoding) {
 630
 631     $body = str_replace("\r\n", "\n", $body);
 632     $encoding = strtolower($encoding);
 633
 634     $encoding_handler = do_hook_function('decode_body', $encoding);
 635
 636
 637     // plugins get first shot at decoding the body
 638     //
 639     if (!empty($encoding_handler) && function_exists($encoding_handler)) {
 640         $body = $encoding_handler('decode', $body);
 641
 642     } elseif ($encoding == 'quoted-printable' ||
 643             $encoding == 'quoted_printable') {
 644         /**
 645          * quoted_printable_decode() function is broken in older
 646          * php versions. Text with \r\n decoding was fixed only
 647          * in php 4.3.0. Minimal code requirement 4.0.4 +
 648          * str_replace("\r\n", "\n", $body); call.
 649          */
 650         $body = quoted_printable_decode($body);
 651     } elseif ($encoding == 'base64') {
 652         $body = base64_decode($body);
 653     }
 654
 655     // All other encodings are returned raw.
 656     return $body;
 657 }
 658
 659 /**
 660  * Decodes headers
 661  *
 662  * This functions decode strings that is encoded according to
 663  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 664  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 665  *
 666  * @param string $string header string that has to be made readable
 667  * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
 668  * @param boolean $htmlsave preserve spaces and sanitize html special characters. defaults to true
 669  * @param boolean $decide decide if string can be utfencoded. defaults to false
 670  * @return string decoded header string
 671  */
 672 function decodeHeader ($string, $utfencode=true,$htmlsave=true,$decide=false) {
 673     global $languages, $squirrelmail_language,$default_charset;
 674     if (is_array($string)) {
 675         $string = implode("\n", $string);
 676     }
 677
 678     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 679             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
 680         $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
 681         // Do we need to return at this point?
 682         // return $string;
 683     }
 684     $i = 0;
 685     $iLastMatch = -2;
 686     $encoded = true;
 687
 688     $aString = explode(' ',$string);
 689     $ret = '';
 690     foreach ($aString as $chunk) {
 691         if ($encoded && $chunk === '') {
 692             continue;
 693         } elseif ($chunk === '') {
 694             $ret .= ' ';
 695             continue;
 696         }
 697         $encoded = false;
 698         /* if encoded words are not separated by a linear-space-white we still catch them */
 699         $j = $i-1;
 700
 701         while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 702             /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
 703             if ($iLastMatch !== $j) {
 704                 if ($htmlsave) {
 705                     $ret .= '&#32;';
 706                 } else {
 707                     $ret .= ' ';
 708                 }
 709             }
 710             $iLastMatch = $i;
 711             $j = $i;
 712             if ($htmlsave) {
 713                 $ret .= htmlspecialchars($res[1]);
 714             } else {
 715                 $ret .= $res[1];
 716             }
 717             $encoding = ucfirst($res[3]);
 718
 719             /* decide about valid decoding */
 720             if ($decide && is_conversion_safe($res[2])) {
 721                 $utfencode=true;
 722                 $can_be_encoded=true;
 723             } else {
 724                 $can_be_encoded=false;
 725             }
 726             switch ($encoding)
 727             {
 728                 case 'B':
 729                     $replace = base64_decode($res[4]);
 730                     if ($utfencode) {
 731                         if ($can_be_encoded) {
 732                             /* convert string to different charset,
 733                              * if functions asks for it (usually in compose)
 734                              */
 735                             $ret .= charset_convert($res[2],$replace,$default_charset);
 736                         } else {
 737                             // convert string to html codes in order to display it
 738                             $ret .= charset_decode($res[2],$replace);
 739                         }
 740                     } else {
 741                         if ($htmlsave) {
 742                             $replace = htmlspecialchars($replace);
 743                         }
 744                         $ret.= $replace;
 745                     }
 746                     break;
 747                 case 'Q':
 748                     $replace = str_replace('_', ' ', $res[4]);
 749                     $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 750                             $replace);
 751                     if ($utfencode) {
 752                         if ($can_be_encoded) {
 753                             /* convert string to different charset,
 754                              * if functions asks for it (usually in compose)
 755                              */
 756                             $replace = charset_convert($res[2], $replace,$default_charset);
 757                         } else {
 758                             // convert string to html codes in order to display it
 759                             $replace = charset_decode($res[2], $replace);
 760                         }
 761                     } else {
 762                         if ($htmlsave) {
 763                             $replace = htmlspecialchars($replace);
 764                         }
 765                     }
 766                     $ret .= $replace;
 767                     break;
 768                 default:
 769                     break;
 770             }
 771             $chunk = $res[5];
 772             $encoded = true;
 773         }
 774         if (!$encoded) {
 775             if ($htmlsave) {
 776                 $ret .= '&#32;';
 777             } else {
 778                 $ret .= ' ';
 779             }
 780         }
 781
 782         if (!$encoded && $htmlsave) {
 783             $ret .= htmlspecialchars($chunk);
 784         } else {
 785             $ret .= $chunk;
 786         }
 787         ++$i;
 788     }
 789     /* remove the first added space */
 790     if ($ret) {
 791         if ($htmlsave) {
 792             $ret = substr($ret,5);
 793         } else {
 794             $ret = substr($ret,1);
 795         }
 796     }
 797
 798     return $ret;
 799 }
 800
 801 /**
 802  * Encodes header
 803  *
 804  * Function uses XTRA_CODE _encodeheader function, if such function exists.
 805  *
 806  * Function uses Q encoding by default and encodes a string according to RFC
 807  * 1522 for use in headers if it contains 8-bit characters or anything that
 808  * looks like it should be encoded.
 809  *
 810  * Function switches to B encoding and encodeHeaderBase64() function, if
 811  * string is 8bit and multibyte character set supported by mbstring extension
 812  * is used. It can cause E_USER_NOTICE errors, if interface is used with
 813  * multibyte character set unsupported by mbstring extension.
 814  *
 815  * @param string $string header string, that has to be encoded
 816  * @return string quoted-printable encoded string
 817  * @todo make $mb_charsets system wide constant
 818  */
 819 function encodeHeader ($string) {
 820     global $default_charset, $languages, $squirrelmail_language;
 821
 822     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 823             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
 824         return  call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
 825     }
 826
 827     // Use B encoding for multibyte charsets
 828     $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
 829     if (in_array($default_charset,$mb_charsets) &&
 830         in_array($default_charset,sq_mb_list_encodings()) &&
 831         sq_is8bit($string)) {
 832         return encodeHeaderBase64($string,$default_charset);
 833     } elseif (in_array($default_charset,$mb_charsets) &&
 834               sq_is8bit($string) &&
 835               ! in_array($default_charset,sq_mb_list_encodings())) {
 836         // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
 837         // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
 838     }
 839
 840     // Encode only if the string contains 8-bit characters or =?
 841     $j = strlen($string);
 842     $max_l = 75 - strlen($default_charset) - 7;
 843     $aRet = array();
 844     $ret = '';
 845     $iEncStart = $enc_init = false;
 846     $cur_l = $iOffset = 0;
 847     for($i = 0; $i < $j; ++$i) {
 848         switch($string{$i})
 849         {
 850             case '=':
 851             case '<':
 852             case '>':
 853             case ',':
 854             case '?':
 855             case '_':
 856                 if ($iEncStart === false) {
 857                     $iEncStart = $i;
 858                 }
 859                 $cur_l+=3;
 860                 if ($cur_l > ($max_l-2)) {
 861                     /* if there is an stringpart that doesn't need encoding, add it */
 862                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 863                     $aRet[] = "=?$default_charset?Q?$ret?=";
 864                     $iOffset = $i;
 865                     $cur_l = 0;
 866                     $ret = '';
 867                     $iEncStart = false;
 868                 } else {
 869                     $ret .= sprintf("=%02X",ord($string{$i}));
 870                 }
 871                 break;
 872             case '(':
 873             case ')':
 874                 if ($iEncStart !== false) {
 875                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 876                     $aRet[] = "=?$default_charset?Q?$ret?=";
 877                     $iOffset = $i;
 878                     $cur_l = 0;
 879                     $ret = '';
 880                     $iEncStart = false;
 881                 }
 882                 break;
 883             case ' ':
 884                 if ($iEncStart !== false) {
 885                     $cur_l++;
 886                     if ($cur_l > $max_l) {
 887                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 888                         $aRet[] = "=?$default_charset?Q?$ret?=";
 889                         $iOffset = $i;
 890                         $cur_l = 0;
 891                         $ret = '';
 892                         $iEncStart = false;
 893                     } else {
 894                         $ret .= '_';
 895                     }
 896                 }
 897                 break;
 898             default:
 899                 $k = ord($string{$i});
 900                 if ($k > 126) {
 901                     if ($iEncStart === false) {
 902                         // do not start encoding in the middle of a string, also take the rest of the word.
 903                         $sLeadString = substr($string,0,$i);
 904                         $aLeadString = explode(' ',$sLeadString);
 905                         $sToBeEncoded = array_pop($aLeadString);
 906                         $iEncStart = $i - strlen($sToBeEncoded);
 907                         $ret .= $sToBeEncoded;
 908                         $cur_l += strlen($sToBeEncoded);
 909                     }
 910                     $cur_l += 3;
 911                     /* first we add the encoded string that reached it's max size */
 912                     if ($cur_l > ($max_l-2)) {
 913                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 914                         $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
 915                         $cur_l = 3;
 916                         $ret = '';
 917                         $iOffset = $i;
 918                         $iEncStart = $i;
 919                     }
 920                     $enc_init = true;
 921                     $ret .= sprintf("=%02X", $k);
 922                 } else {
 923                     if ($iEncStart !== false) {
 924                         $cur_l++;
 925                         if ($cur_l > $max_l) {
 926                             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 927                             $aRet[] = "=?$default_charset?Q?$ret?=";
 928                             $iEncStart = false;
 929                             $iOffset = $i;
 930                             $cur_l = 0;
 931                             $ret = '';
 932                         } else {
 933                             $ret .= $string{$i};
 934                         }
 935                     }
 936                 }
 937                 break;
 938         }
 939     }
 940
 941     if ($enc_init) {
 942         if ($iEncStart !== false) {
 943             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 944             $aRet[] = "=?$default_charset?Q?$ret?=";
 945         } else {
 946             $aRet[] = substr($string,$iOffset);
 947         }
 948         $string = implode('',$aRet);
 949     }
 950     return $string;
 951 }
 952
 953 /**
 954  * Encodes string according to rfc2047 B encoding header formating rules
 955  *
 956  * It is recommended way to encode headers with character sets that store
 957  * symbols in more than one byte.
 958  *
 959  * Function requires mbstring support. If required mbstring functions are missing,
 960  * function returns false and sets E_USER_WARNING level error message.
 961  *
 962  * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
 963  * that mbstring functions will generate E_WARNING errors, if unsupported
 964  * character set is used. mb_encode_mimeheader function provided by php
 965  * mbstring extension is not used in order to get better control of header
 966  * encoding.
 967  *
 968  * Used php code functions - function_exists(), trigger_error(), strlen()
 969  * (is used with charset names and base64 strings). Used php mbstring
 970  * functions - mb_strlen and mb_substr.
 971  *
 972  * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
 973  * encoding), rfc 2822 (header folding)
 974  *
 975  * @param string $string header string that must be encoded
 976  * @param string $charset character set. Must be supported by mbstring extension.
 977  * Use sq_mb_list_encodings() to detect supported charsets.
 978  * @return string string encoded according to rfc2047 B encoding formating rules
 979  * @since 1.5.1
 980  * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
 981  * @todo Do we want to control max length of header?
 982  * @todo Do we want to control EOL (end-of-line) marker?
 983  * @todo Do we want to translate error message?
 984  */
 985 function encodeHeaderBase64($string,$charset) {
 986     /**
 987      * Check mbstring function requirements.
 988      */
 989     if (! function_exists('mb_strlen') ||
 990         ! function_exists('mb_substr')) {
 991         // set E_USER_WARNING
 992         trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
 993         // return false
 994         return false;
 995     }
 996
 997     // initial return array
 998     $aRet = array();
 999
1000     /**
1001      * header length = 75 symbols max (same as in encodeHeader)
1002      * remove $charset length
1003      * remove =? ? ?= (5 chars)
1004      * remove 2 more chars (\r\n ?)
1005      */
1006     $iMaxLength = 75 - strlen($charset) - 7;
1007
1008     // set first character position
1009     $iStartCharNum = 0;
1010
1011     // loop through all characters. count characters and not bytes.
1012     for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1013         // encode string from starting character to current character.
1014         $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1015
1016         // Check encoded string length
1017         if(strlen($encoded_string)>$iMaxLength) {
1018             // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1019             $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1020
1021             // set new starting character
1022             $iStartCharNum = $iCharNum-1;
1023
1024             // encode last char (in case it is last character in string)
1025             $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1026         } // if string is shorter than max length - add next character
1027     }
1028
1029     // add last encoded string to array
1030     $aRet[] = $encoded_string;
1031
1032     // set initial return string
1033     $sRet = '';
1034
1035     // loop through encoded strings
1036     foreach($aRet as $string) {
1037         // TODO: Do we want to control EOL (end-of-line) marker
1038         if ($sRet!='') $sRet.= " ";
1039
1040         // add header tags and encoded string to return string
1041         $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1042     }
1043
1044     return $sRet;
1045 }
1046
1047 /* This function trys to locate the entity_id of a specific mime element */
1048 function find_ent_id($id, $message) {
1049     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1050         if ($message->entities[$i]->header->type0 == 'multipart')  {
1051             $ret = find_ent_id($id, $message->entities[$i]);
1052         } else {
1053             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
1054 //                if (sq_check_save_extension($message->entities[$i])) {
1055                 return $message->entities[$i]->entity_id;
1056 //                }
1057             } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1058                 /**
1059                  * This is part of a fix for Outlook Express 6.x generating
1060                  * cid URLs without creating content-id headers
1061                  * @@JA - 20050207
1062                  */
1063                 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1064                     return $message->entities[$i]->entity_id;
1065                 }
1066             }
1067         }
1068     }
1069     return $ret;
1070 }
1071
1072 function sq_check_save_extension($message) {
1073     $filename = $message->getFilename();
1074     $ext = substr($filename, strrpos($filename,'.')+1);
1075     $save_extensions = array('jpg','jpeg','gif','png','bmp');
1076     return in_array($ext, $save_extensions);
1077 }
1078
1079
1080 /**
1081  ** HTMLFILTER ROUTINES
1082  */
1083
1084 /**
1085  * This function checks attribute values for entity-encoded values
1086  * and returns them translated into 8-bit strings so we can run
1087  * checks on them.
1088  *
1089  * @param  $attvalue A string to run entity check against.
1090  * @return           Nothing, modifies a reference value.
1091  */
1092 function sq_defang(&$attvalue){
1093     $me = 'sq_defang';
1094     /**
1095      * Skip this if there aren't ampersands or backslashes.
1096      */
1097     if (strpos($attvalue, '&') === false
1098         && strpos($attvalue, '\\') === false){
1099         return;
1100     }
1101     $m = false;
1102     do {
1103         $m = false;
1104         $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1105         $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1106         $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1107     } while ($m == true);
1108     $attvalue = stripslashes($attvalue);
1109 }
1110
1111 /**
1112  * Kill any tabs, newlines, or carriage returns. Our friends the
1113  * makers of the browser with 95% market value decided that it'd
1114  * be funny to make "java[tab]script" be just as good as "javascript".
1115  *
1116  * @param  attvalue  The attribute value before extraneous spaces removed.
1117  * @return attvalue  Nothing, modifies a reference value.
1118  */
1119 function sq_unspace(&$attvalue){
1120     $me = 'sq_unspace';
1121     if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1122         $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1123                                 Array('',   '',   '',   '',   ''), $attvalue);
1124     }
1125 }
1126
1127 /**
1128  * This function returns the final tag out of the tag name, an array
1129  * of attributes, and the type of the tag. This function is called by
1130  * sq_sanitize internally.
1131  *
1132  * @param  $tagname  the name of the tag.
1133  * @param  $attary   the array of attributes and their values
1134  * @param  $tagtype  The type of the tag (see in comments).
1135  * @return           a string with the final tag representation.
1136  */
1137 function sq_tagprint($tagname, $attary, $tagtype){
1138     $me = 'sq_tagprint';
1139
1140     if ($tagtype == 2){
1141         $fulltag = '</' . $tagname . '>';
1142     } else {
1143         $fulltag = '<' . $tagname;
1144         if (is_array($attary) && sizeof($attary)){
1145             $atts = Array();
1146             while (list($attname, $attvalue) = each($attary)){
1147                 array_push($atts, "$attname=$attvalue");
1148             }
1149             $fulltag .= ' ' . join(" ", $atts);
1150         }
1151         if ($tagtype == 3){
1152             $fulltag .= ' /';
1153         }
1154         $fulltag .= '>';
1155     }
1156     return $fulltag;
1157 }
1158
1159 /**
1160  * A small helper function to use with array_walk. Modifies a by-ref
1161  * value and makes it lowercase.
1162  *
1163  * @param  $val a value passed by-ref.
1164  * @return      void since it modifies a by-ref value.
1165  */
1166 function sq_casenormalize(&$val){
1167     $val = strtolower($val);
1168 }
1169
1170 /**
1171  * This function skips any whitespace from the current position within
1172  * a string and to the next non-whitespace value.
1173  *
1174  * @param  $body   the string
1175  * @param  $offset the offset within the string where we should start
1176  *                 looking for the next non-whitespace character.
1177  * @return         the location within the $body where the next
1178  *                 non-whitespace char is located.
1179  */
1180 function sq_skipspace($body, $offset){
1181     $me = 'sq_skipspace';
1182     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1183     if (sizeof($matches{1})){
1184         $count = strlen($matches{1});
1185         $offset += $count;
1186     }
1187     return $offset;
1188 }
1189
1190 /**
1191  * This function looks for the next character within a string.  It's
1192  * really just a glorified "strpos", except it catches if failures
1193  * nicely.
1194  *
1195  * @param  $body   The string to look for needle in.
1196  * @param  $offset Start looking from this position.
1197  * @param  $needle The character/string to look for.
1198  * @return         location of the next occurance of the needle, or
1199  *                 strlen($body) if needle wasn't found.
1200  */
1201 function sq_findnxstr($body, $offset, $needle){
1202     $me  = 'sq_findnxstr';
1203     $pos = strpos($body, $needle, $offset);
1204     if ($pos === FALSE){
1205         $pos = strlen($body);
1206     }
1207     return $pos;
1208 }
1209
1210 /**
1211  * This function takes a PCRE-style regexp and tries to match it
1212  * within the string.
1213  *
1214  * @param  $body   The string to look for needle in.
1215  * @param  $offset Start looking from here.
1216  * @param  $reg    A PCRE-style regex to match.
1217  * @return         Returns a false if no matches found, or an array
1218  *                 with the following members:
1219  *                 - integer with the location of the match within $body
1220  *                 - string with whatever content between offset and the match
1221  *                 - string with whatever it is we matched
1222  */
1223 function sq_findnxreg($body, $offset, $reg){
1224     $me = 'sq_findnxreg';
1225     $matches = Array();
1226     $retarr = Array();
1227     preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1228     if (!isset($matches{0}) || !$matches{0}){
1229         $retarr = false;
1230     } else {
1231         $retarr{0} = $offset + strlen($matches{1});
1232         $retarr{1} = $matches{1};
1233         $retarr{2} = $matches{2};
1234     }
1235     return $retarr;
1236 }
1237
1238 /**
1239  * This function looks for the next tag.
1240  *
1241  * @param  $body   String where to look for the next tag.
1242  * @param  $offset Start looking from here.
1243  * @return         false if no more tags exist in the body, or
1244  *                 an array with the following members:
1245  *                 - string with the name of the tag
1246  *                 - array with attributes and their values
1247  *                 - integer with tag type (1, 2, or 3)
1248  *                 - integer where the tag starts (starting "<")
1249  *                 - integer where the tag ends (ending ">")
1250  *                 first three members will be false, if the tag is invalid.
1251  */
1252 function sq_getnxtag($body, $offset){
1253     $me = 'sq_getnxtag';
1254     if ($offset > strlen($body)){
1255         return false;
1256     }
1257     $lt = sq_findnxstr($body, $offset, "<");
1258     if ($lt == strlen($body)){
1259         return false;
1260     }
1261     /**
1262      * We are here:
1263      * blah blah <tag attribute="value">
1264      * \---------^
1265      */
1266     $pos = sq_skipspace($body, $lt+1);
1267     if ($pos >= strlen($body)){
1268         return Array(false, false, false, $lt, strlen($body));
1269     }
1270     /**
1271      * There are 3 kinds of tags:
1272      * 1. Opening tag, e.g.:
1273      *    <a href="blah">
1274      * 2. Closing tag, e.g.:
1275      *    </a>
1276      * 3. XHTML-style content-less tag, e.g.:
1277      *    <img src="blah" />
1278      */
1279     $tagtype = false;
1280     switch (substr($body, $pos, 1)){
1281         case '/':
1282             $tagtype = 2;
1283             $pos++;
1284             break;
1285         case '!':
1286             /**
1287              * A comment or an SGML declaration.
1288              */
1289             if (substr($body, $pos+1, 2) == "--"){
1290                 $gt = strpos($body, "-->", $pos);
1291                 if ($gt === false){
1292                     $gt = strlen($body);
1293                 } else {
1294                     $gt += 2;
1295                 }
1296                 return Array(false, false, false, $lt, $gt);
1297             } else {
1298                 $gt = sq_findnxstr($body, $pos, ">");
1299                 return Array(false, false, false, $lt, $gt);
1300             }
1301             break;
1302         default:
1303             /**
1304              * Assume tagtype 1 for now. If it's type 3, we'll switch values
1305              * later.
1306              */
1307             $tagtype = 1;
1308             break;
1309     }
1310
1311     $tag_start = $pos;
1312     $tagname = '';
1313     /**
1314      * Look for next [\W-_], which will indicate the end of the tag name.
1315      */
1316     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1317     if ($regary == false){
1318         return Array(false, false, false, $lt, strlen($body));
1319     }
1320     list($pos, $tagname, $match) = $regary;
1321     $tagname = strtolower($tagname);
1322
1323     /**
1324      * $match can be either of these:
1325      * '>'  indicating the end of the tag entirely.
1326      * '\s' indicating the end of the tag name.
1327      * '/'  indicating that this is type-3 xhtml tag.
1328      *
1329      * Whatever else we find there indicates an invalid tag.
1330      */
1331     switch ($match){
1332         case '/':
1333             /**
1334              * This is an xhtml-style tag with a closing / at the
1335              * end, like so: <img src="blah" />. Check if it's followed
1336              * by the closing bracket. If not, then this tag is invalid
1337              */
1338             if (substr($body, $pos, 2) == "/>"){
1339                 $pos++;
1340                 $tagtype = 3;
1341             } else {
1342                 $gt = sq_findnxstr($body, $pos, ">");
1343                 $retary = Array(false, false, false, $lt, $gt);
1344                 return $retary;
1345             }
1346         case '>':
1347             return Array($tagname, false, $tagtype, $lt, $pos);
1348             break;
1349         default:
1350             /**
1351              * Check if it's whitespace
1352              */
1353             if (!preg_match('/\s/', $match)){
1354                 /**
1355                  * This is an invalid tag! Look for the next closing ">".
1356                  */
1357                 $gt = sq_findnxstr($body, $lt, ">");
1358                 return Array(false, false, false, $lt, $gt);
1359             }
1360             break;
1361     }
1362
1363     /**
1364      * At this point we're here:
1365      * <tagname  attribute='blah'>
1366      * \-------^
1367      *
1368      * At this point we loop in order to find all attributes.
1369      */
1370     $attname = '';
1371     $atttype = false;
1372     $attary = Array();
1373
1374     while ($pos <= strlen($body)){
1375         $pos = sq_skipspace($body, $pos);
1376         if ($pos == strlen($body)){
1377             /**
1378              * Non-closed tag.
1379              */
1380             return Array(false, false, false, $lt, $pos);
1381         }
1382         /**
1383          * See if we arrived at a ">" or "/>", which means that we reached
1384          * the end of the tag.
1385          */
1386         $matches = Array();
1387         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1388             /**
1389              * Yep. So we did.
1390              */
1391             $pos += strlen($matches{1});
1392             if ($matches{2} == "/>"){
1393                 $tagtype = 3;
1394                 $pos++;
1395             }
1396             return Array($tagname, $attary, $tagtype, $lt, $pos);
1397         }
1398
1399         /**
1400          * There are several types of attributes, with optional
1401          * [:space:] between members.
1402          * Type 1:
1403          *   attrname[:space:]=[:space:]'CDATA'
1404          * Type 2:
1405          *   attrname[:space:]=[:space:]"CDATA"
1406          * Type 3:
1407          *   attr[:space:]=[:space:]CDATA
1408          * Type 4:
1409          *   attrname
1410          *
1411          * We leave types 1 and 2 the same, type 3 we check for
1412          * '"' and convert to "&quot" if needed, then wrap in
1413          * double quotes. Type 4 we convert into:
1414          * attrname="yes".
1415          */
1416         $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1417         if ($regary == false){
1418             /**
1419              * Looks like body ended before the end of tag.
1420              */
1421             return Array(false, false, false, $lt, strlen($body));
1422         }
1423         list($pos, $attname, $match) = $regary;
1424         $attname = strtolower($attname);
1425         /**
1426          * We arrived at the end of attribute name. Several things possible
1427          * here:
1428          * '>'  means the end of the tag and this is attribute type 4
1429          * '/'  if followed by '>' means the same thing as above
1430          * '\s' means a lot of things -- look what it's followed by.
1431          *      anything else means the attribute is invalid.
1432          */
1433         switch($match){
1434             case '/':
1435                 /**
1436                  * This is an xhtml-style tag with a closing / at the
1437                  * end, like so: <img src="blah" />. Check if it's followed
1438                  * by the closing bracket. If not, then this tag is invalid
1439                  */
1440                 if (substr($body, $pos, 2) == "/>"){
1441                     $pos++;
1442                     $tagtype = 3;
1443                 } else {
1444                     $gt = sq_findnxstr($body, $pos, ">");
1445                     $retary = Array(false, false, false, $lt, $gt);
1446                     return $retary;
1447                 }
1448             case '>':
1449                 $attary{$attname} = '"yes"';
1450                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1451                 break;
1452             default:
1453                 /**
1454                  * Skip whitespace and see what we arrive at.
1455                  */
1456                 $pos = sq_skipspace($body, $pos);
1457                 $char = substr($body, $pos, 1);
1458                 /**
1459                  * Two things are valid here:
1460                  * '=' means this is attribute type 1 2 or 3.
1461                  * \w means this was attribute type 4.
1462                  * anything else we ignore and re-loop. End of tag and
1463                  * invalid stuff will be caught by our checks at the beginning
1464                  * of the loop.
1465                  */
1466                 if ($char == "="){
1467                     $pos++;
1468                     $pos = sq_skipspace($body, $pos);
1469                     /**
1470                      * Here are 3 possibilities:
1471                      * "'"  attribute type 1
1472                      * '"'  attribute type 2
1473                      * everything else is the content of tag type 3
1474                      */
1475                     $quot = substr($body, $pos, 1);
1476                     if ($quot == "'"){
1477                         $regary = sq_findnxreg($body, $pos+1, "\'");
1478                         if ($regary == false){
1479                             return Array(false, false, false, $lt, strlen($body));
1480                         }
1481                         list($pos, $attval, $match) = $regary;
1482                         $pos++;
1483                         $attary{$attname} = "'" . $attval . "'";
1484                     } else if ($quot == '"'){
1485                         $regary = sq_findnxreg($body, $pos+1, '\"');
1486                         if ($regary == false){
1487                             return Array(false, false, false, $lt, strlen($body));
1488                         }
1489                         list($pos, $attval, $match) = $regary;
1490                         $pos++;
1491                         $attary{$attname} = '"' . $attval . '"';
1492                     } else {
1493                         /**
1494                          * These are hateful. Look for \s, or >.
1495                          */
1496                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1497                         if ($regary == false){
1498                             return Array(false, false, false, $lt, strlen($body));
1499                         }
1500                         list($pos, $attval, $match) = $regary;
1501                         /**
1502                          * If it's ">" it will be caught at the top.
1503                          */
1504                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1505                         $attary{$attname} = '"' . $attval . '"';
1506                     }
1507                 } else if (preg_match("|[\w/>]|", $char)) {
1508                     /**
1509                      * That was attribute type 4.
1510                      */
1511                     $attary{$attname} = '"yes"';
1512                 } else {
1513                     /**
1514                      * An illegal character. Find next '>' and return.
1515                      */
1516                     $gt = sq_findnxstr($body, $pos, ">");
1517                     return Array(false, false, false, $lt, $gt);
1518                 }
1519                 break;
1520         }
1521     }
1522     /**
1523      * The fact that we got here indicates that the tag end was never
1524      * found. Return invalid tag indication so it gets stripped.
1525      */
1526     return Array(false, false, false, $lt, strlen($body));
1527 }
1528
1529 /**
1530  * Translates entities into literal values so they can be checked.
1531  *
1532  * @param $attvalue the by-ref value to check.
1533  * @param $regex    the regular expression to check against.
1534  * @param $hex      whether the entites are hexadecimal.
1535  * @return          True or False depending on whether there were matches.
1536  */
1537 function sq_deent(&$attvalue, $regex, $hex=false){
1538     $me = 'sq_deent';
1539     $ret_match = false;
1540     preg_match_all($regex, $attvalue, $matches);
1541     if (is_array($matches) && sizeof($matches[0]) > 0){
1542         $repl = Array();
1543         for ($i = 0; $i < sizeof($matches[0]); $i++){
1544             $numval = $matches[1][$i];
1545             if ($hex){
1546                 $numval = hexdec($numval);
1547             }
1548             $repl{$matches[0][$i]} = chr($numval);
1549         }
1550         $attvalue = strtr($attvalue, $repl);
1551         return true;
1552     } else {
1553         return false;
1554     }
1555 }
1556
1557 /**
1558  * This function runs various checks against the attributes.
1559  *
1560  * @param  $tagname         String with the name of the tag.
1561  * @param  $attary          Array with all tag attributes.
1562  * @param  $rm_attnames     See description for sq_sanitize
1563  * @param  $bad_attvals     See description for sq_sanitize
1564  * @param  $add_attr_to_tag See description for sq_sanitize
1565  * @param  $message         message object
1566  * @param  $id              message id
1567  * @return                  Array with modified attributes.
1568  */
1569 function sq_fixatts($tagname,
1570                     $attary,
1571                     $rm_attnames,
1572                     $bad_attvals,
1573                     $add_attr_to_tag,
1574                     $message,
1575                     $id,
1576                     $mailbox
1577                     ){
1578     $me = 'sq_fixatts';
1579     while (list($attname, $attvalue) = each($attary)){
1580         /**
1581          * See if this attribute should be removed.
1582          */
1583         foreach ($rm_attnames as $matchtag=>$matchattrs){
1584             if (preg_match($matchtag, $tagname)){
1585                 foreach ($matchattrs as $matchattr){
1586                     if (preg_match($matchattr, $attname)){
1587                         unset($attary{$attname});
1588                         continue;
1589                     }
1590                 }
1591             }
1592         }
1593         /**
1594          * Remove any backslashes, entities, and extraneous whitespace.
1595          */
1596         sq_defang($attvalue);
1597         sq_unspace($attvalue);
1598
1599         /**
1600          * Now let's run checks on the attvalues.
1601          * I don't expect anyone to comprehend this. If you do,
1602          * get in touch with me so I can drive to where you live and
1603          * shake your hand personally. :)
1604          */
1605         foreach ($bad_attvals as $matchtag=>$matchattrs){
1606             if (preg_match($matchtag, $tagname)){
1607                 foreach ($matchattrs as $matchattr=>$valary){
1608                     if (preg_match($matchattr, $attname)){
1609                         /**
1610                          * There are two arrays in valary.
1611                          * First is matches.
1612                          * Second one is replacements
1613                          */
1614                         list($valmatch, $valrepl) = $valary;
1615                         $newvalue =
1616                             preg_replace($valmatch, $valrepl, $attvalue);
1617                         if ($newvalue != $attvalue){
1618                             $attary{$attname} = $newvalue;
1619                         }
1620                     }
1621                 }
1622             }
1623         }
1624
1625         /**
1626          * Replace empty src tags with the blank image.  src is only used
1627          * for frames, images, and image inputs.  Doing a replace should
1628          * not affect them working as should be, however it will stop
1629          * IE from being kicked off when src for img tags are not set
1630          */
1631         if (($attname == 'src') && ($attvalue == '""')) {
1632             $attary{$attname} = '"' . SM_PATH . 'images/blank.png"';
1633         }
1634
1635         /**
1636          * Turn cid: urls into http-friendly ones.
1637          */
1638         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1639             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1640         }
1641
1642         /**
1643          * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1644          * One day MS might actually make it match something useful, for now, falling
1645          * back to using cid2http, so we can grab the blank.png.
1646          */
1647         if (preg_match("/^[\'\"]\s*outbind:\/\//si", $attvalue)) {
1648             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1649         }
1650
1651     }
1652     /**
1653      * See if we need to append any attributes to this tag.
1654      */
1655     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1656         if (preg_match($matchtag, $tagname)){
1657             $attary = array_merge($attary, $addattary);
1658         }
1659     }
1660     return $attary;
1661 }
1662
1663 /**
1664  * This function edits the style definition to make them friendly and
1665  * usable in SquirrelMail.
1666  *
1667  * @param  $message  the message object
1668  * @param  $id       the message id
1669  * @param  $content  a string with whatever is between <style> and </style>
1670  * @param  $mailbox  the message mailbox
1671  * @return           a string with edited content.
1672  */
1673 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1674     global $view_unsafe_images;
1675     $me = 'sq_fixstyle';
1676     $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1677     if ($ret == FALSE){
1678         return array(FALSE, strlen($body));
1679     }
1680     $newpos = $ret[0] + strlen($ret[2]);
1681     $content = $ret[1];
1682     /**
1683     * First look for general BODY style declaration, which would be
1684     * like so:
1685     * body {background: blah-blah}
1686     * and change it to .bodyclass so we can just assign it to a <div>
1687     */
1688     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1689     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1690     /**
1691     * Fix url('blah') declarations.
1692     */
1693     //   $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1694     //                           "url(\\1$secremoveimg\\2)", $content);
1695     // remove NUL
1696     $content = str_replace("\0", "", $content);
1697
1698     // NB I insert NUL characters to keep to avoid an infinite loop. They are removed after the loop.
1699     while (preg_match("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si", $content, $matches)) {
1700         $sProto = strtolower($matches[1]);
1701         switch ($sProto) {
1702             /**
1703              * Fix url('https*://.*) declarations but only if $view_unsafe_images
1704              * is false.
1705              */
1706             case 'https':
1707             case 'http':
1708                 if (!$view_unsafe_images){
1709
1710                     $sExpr = "/url\s*\(\s*[\'\"]?\s*$sProto*:.*[\'\"]?\s*\)/si";
1711                     $content = preg_replace($sExpr, "u\0r\0l(\\1$secremoveimg\\2)", $content);
1712
1713                 } else {
1714                     $content = preg_replace('/url/i',"u\0r\0l",$content);
1715                 }
1716                 break;
1717             /**
1718              * Fix urls that refer to cid:
1719              */
1720             case 'cid':
1721                 $cidurl = 'cid:'. $matches[2];
1722                 $httpurl = sq_cid2http($message, $id, $cidurl, $mailbox);
1723                 // escape parentheses that can modify the regular expression
1724                 $cidurl = str_replace(array('(',')'),array('\\(','\\)'),$cidurl);
1725                 $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1726                                         "u\0r\0l($httpurl)", $content);
1727                 break;
1728             default:
1729                 /**
1730                  * replace url with protocol other then the white list
1731                  * http,https and cid by an empty string.
1732                  */
1733                 $content = preg_replace("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si",
1734                                  "", $content);
1735                 break;
1736         }
1737     }
1738     // remove NUL
1739     $content = str_replace("\0", "", $content);
1740    /**
1741     * Remove any backslashes, entities, and extraneous whitespace.
1742     */
1743     $contentTemp = $content;
1744     sq_defang($contentTemp);
1745     sq_unspace($contentTemp);
1746
1747     /**
1748      * Fix stupid css declarations which lead to vulnerabilities
1749      * in IE.
1750      */
1751     $match   = Array('/\/\*.*\*\//',
1752                     '/expression/i',
1753                     '/behaviou*r/i',
1754                     '/binding/i',
1755                     '/include-source/i');
1756     $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy');
1757     $contentNew = preg_replace($match, $replace, $contentTemp);
1758     if ($contentNew !== $contentTemp) {
1759         // insecure css declarations are used. From now on we don't care
1760         // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
1761         $content = $contentNew;
1762     }
1763     return array($content, $newpos);
1764 }
1765
1766
1767 /**
1768  * This function converts cid: url's into the ones that can be viewed in
1769  * the browser.
1770  *
1771  * @param  $message  the message object
1772  * @param  $id       the message id
1773  * @param  $cidurl   the cid: url.
1774  * @param  $mailbox  the message mailbox
1775  * @return           a string with a http-friendly url
1776  */
1777 function sq_cid2http($message, $id, $cidurl, $mailbox){
1778     /**
1779      * Get rid of quotes.
1780      */
1781     $quotchar = substr($cidurl, 0, 1);
1782     if ($quotchar == '"' || $quotchar == "'"){
1783         $cidurl = str_replace($quotchar, "", $cidurl);
1784     } else {
1785         $quotchar = '';
1786     }
1787     $cidurl = substr(trim($cidurl), 4);
1788
1789     $match_str = '/\{.*?\}\//';
1790     $str_rep = '';
1791     $cidurl = preg_replace($match_str, $str_rep, $cidurl);
1792
1793     $linkurl = find_ent_id($cidurl, $message);
1794     /* in case of non-save cid links $httpurl should be replaced by a sort of
1795        unsave link image */
1796     $httpurl = '';
1797
1798     /**
1799      * This is part of a fix for Outlook Express 6.x generating
1800      * cid URLs without creating content-id headers. These images are
1801      * not part of the multipart/related html mail. The html contains
1802      * <img src="cid:{some_id}/image_filename.ext"> references to
1803      * attached images with as goal to render them inline although
1804      * the attachment disposition property is not inline.
1805      */
1806
1807     if (empty($linkurl)) {
1808         if (preg_match('/{.*}\//', $cidurl)) {
1809             $cidurl = preg_replace('/{.*}\//','', $cidurl);
1810             if (!empty($cidurl)) {
1811                 $linkurl = find_ent_id($cidurl, $message);
1812             }
1813         }
1814     }
1815
1816     if (!empty($linkurl)) {
1817         $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1818             "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1819             '&amp;ent_id=' . $linkurl . $quotchar;
1820     } else {
1821         /**
1822          * If we couldn't generate a proper img url, drop in a blank image
1823          * instead of sending back empty, otherwise it causes unusual behaviour
1824          */
1825         $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
1826     }
1827
1828     return $httpurl;
1829 }
1830
1831 /**
1832  * This function changes the <body> tag into a <div> tag since we
1833  * can't really have a body-within-body.
1834  *
1835  * @param  $attary   an array of attributes and values of <body>
1836  * @param  $mailbox  mailbox we're currently reading (for cid2http)
1837  * @param  $message  current message (for cid2http)
1838  * @param  $id       current message id (for cid2http)
1839  * @return           a modified array of attributes to be set for <div>
1840  */
1841 function sq_body2div($attary, $mailbox, $message, $id){
1842     $me = 'sq_body2div';
1843     $divattary = Array('class' => "'bodyclass'");
1844     $text = '#000000';
1845     $has_bgc_stl = $has_txt_stl = false;
1846     $styledef = '';
1847     if (is_array($attary) && sizeof($attary) > 0){
1848         foreach ($attary as $attname=>$attvalue){
1849             $quotchar = substr($attvalue, 0, 1);
1850             $attvalue = str_replace($quotchar, "", $attvalue);
1851             switch ($attname){
1852                 case 'background':
1853                     $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1854                     $styledef .= "background-image: url('$attvalue'); ";
1855                     break;
1856                 case 'bgcolor':
1857                     $has_bgc_stl = true;
1858                     $styledef .= "background-color: $attvalue; ";
1859                     break;
1860                 case 'text':
1861                     $has_txt_stl = true;
1862                     $styledef .= "color: $attvalue; ";
1863                     break;
1864             }
1865         }
1866         // Outlook defines a white bgcolor and no text color. This can lead to
1867         // white text on a white bg with certain themes.
1868         if ($has_bgc_stl && !$has_txt_stl) {
1869             $styledef .= "color: $text; ";
1870         }
1871         if (strlen($styledef) > 0){
1872             $divattary{"style"} = "\"$styledef\"";
1873         }
1874     }
1875     return $divattary;
1876 }
1877
1878 /**
1879  * This is the main function and the one you should actually be calling.
1880  * There are several variables you should be aware of an which need
1881  * special description.
1882  *
1883  * Since the description is quite lengthy, see it here:
1884  * http://linux.duke.edu/projects/mini/htmlfilter/
1885  *
1886  * @param $body                 the string with HTML you wish to filter
1887  * @param $tag_list             see description above
1888  * @param $rm_tags_with_content see description above
1889  * @param $self_closing_tags    see description above
1890  * @param $force_tag_closing    see description above
1891  * @param $rm_attnames          see description above
1892  * @param $bad_attvals          see description above
1893  * @param $add_attr_to_tag      see description above
1894  * @param $message              message object
1895  * @param $id                   message id
1896  * @return                      sanitized html safe to show on your pages.
1897  */
1898 function sq_sanitize($body,
1899                      $tag_list,
1900                      $rm_tags_with_content,
1901                      $self_closing_tags,
1902                      $force_tag_closing,
1903                      $rm_attnames,
1904                      $bad_attvals,
1905                      $add_attr_to_tag,
1906                      $message,
1907                      $id,
1908                      $mailbox
1909                      ){
1910     $me = 'sq_sanitize';
1911     $rm_tags = array_shift($tag_list);
1912     /**
1913      * Normalize rm_tags and rm_tags_with_content.
1914      */
1915     @array_walk($tag_list, 'sq_casenormalize');
1916     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1917     @array_walk($self_closing_tags, 'sq_casenormalize');
1918     /**
1919      * See if tag_list is of tags to remove or tags to allow.
1920      * false  means remove these tags
1921      * true   means allow these tags
1922      */
1923     $curpos = 0;
1924     $open_tags = Array();
1925     $trusted = "\n<!-- begin sanitized html -->\n";
1926     $skip_content = false;
1927     /**
1928      * Take care of netscape's stupid javascript entities like
1929      * &{alert('boo')};
1930      */
1931     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1932
1933     while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1934         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1935         $free_content = substr($body, $curpos, $lt-$curpos);
1936         /**
1937          * Take care of <style>
1938          */
1939         if ($tagname == "style" && $tagtype == 1){
1940             list($free_content, $curpos) =
1941                 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
1942             if ($free_content != FALSE){
1943                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1944                 $trusted .= $free_content;
1945                 $trusted .= sq_tagprint($tagname, false, 2);
1946             }
1947             continue;
1948         }
1949         if ($skip_content == false){
1950             $trusted .= $free_content;
1951         }
1952         if ($tagname != FALSE){
1953             if ($tagtype == 2){
1954                 if ($skip_content == $tagname){
1955                     /**
1956                      * Got to the end of tag we needed to remove.
1957                      */
1958                     $tagname = false;
1959                     $skip_content = false;
1960                 } else {
1961                     if ($skip_content == false){
1962                         if ($tagname == "body"){
1963                             $tagname = "div";
1964                         }
1965                         if (isset($open_tags{$tagname}) &&
1966                                 $open_tags{$tagname} > 0){
1967                             $open_tags{$tagname}--;
1968                         } else {
1969                             $tagname = false;
1970                         }
1971                     }
1972                 }
1973             } else {
1974                 /**
1975                  * $rm_tags_with_content
1976                  */
1977                 if ($skip_content == false){
1978                     /**
1979                      * See if this is a self-closing type and change
1980                      * tagtype appropriately.
1981                      */
1982                     if ($tagtype == 1
1983                             && in_array($tagname, $self_closing_tags)){
1984                         $tagtype = 3;
1985                     }
1986                     /**
1987                      * See if we should skip this tag and any content
1988                      * inside it.
1989                      */
1990                     if ($tagtype == 1 &&
1991                             in_array($tagname, $rm_tags_with_content)){
1992                         $skip_content = $tagname;
1993                     } else {
1994                         if (($rm_tags == false
1995                                     && in_array($tagname, $tag_list)) ||
1996                                 ($rm_tags == true &&
1997                                  !in_array($tagname, $tag_list))){
1998                             $tagname = false;
1999                         } else {
2000                             /**
2001                              * Convert body into div.
2002                              */
2003                             if ($tagname == "body"){
2004                                 $tagname = "div";
2005                                 $attary = sq_body2div($attary, $mailbox,
2006                                         $message, $id);
2007                             }
2008                             if ($tagtype == 1){
2009                                 if (isset($open_tags{$tagname})){
2010                                     $open_tags{$tagname}++;
2011                                 } else {
2012                                     $open_tags{$tagname}=1;
2013                                 }
2014                             }
2015                             /**
2016                              * This is where we run other checks.
2017                              */
2018                             if (is_array($attary) && sizeof($attary) > 0){
2019                                 $attary = sq_fixatts($tagname,
2020                                                      $attary,
2021                                                      $rm_attnames,
2022                                                      $bad_attvals,
2023                                                      $add_attr_to_tag,
2024                                                      $message,
2025                                                      $id,
2026                                                      $mailbox
2027                                                      );
2028                             }
2029                         }
2030                     }
2031                 }
2032             }
2033             if ($tagname != false && $skip_content == false){
2034                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2035             }
2036         }
2037         $curpos = $gt+1;
2038     }
2039     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2040     if ($force_tag_closing == true){
2041         foreach ($open_tags as $tagname=>$opentimes){
2042             while ($opentimes > 0){
2043                 $trusted .= '</' . $tagname . '>';
2044                 $opentimes--;
2045             }
2046         }
2047         $trusted .= "\n";
2048     }
2049     $trusted .= "<!-- end sanitized html -->\n";
2050     return $trusted;
2051 }
2052
2053 /**
2054  * This is a wrapper function to call html sanitizing routines.
2055  *
2056  * @param  $body  the body of the message
2057  * @param  $id    the id of the message
2058
2059  * @param  $message
2060  * @param  $mailbox
2061  * @param  boolean $take_mailto_links When TRUE, converts mailto: links
2062  *                                    into internal SM compose links
2063  *                                    (optional; default = TRUE)
2064  * @return        a string with html safe to display in the browser.
2065  */
2066 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2067
2068     require_once(SM_PATH . 'functions/url_parser.php');  // for $MailTo_PReg_Match
2069
2070     global $attachment_common_show_images, $view_unsafe_images,
2071            $has_unsafe_images;
2072     /**
2073      * Don't display attached images in HTML mode.
2074      */
2075     $attachment_common_show_images = false;
2076     $tag_list = Array(
2077             false,
2078             "object",
2079             "meta",
2080             "html",
2081             "head",
2082             "base",
2083             "link",
2084             "frame",
2085             "iframe",
2086             "plaintext",
2087             "marquee"
2088             );
2089
2090     $rm_tags_with_content = Array(
2091             "script",
2092             "applet",
2093             "embed",
2094             "title",
2095             "frameset",
2096             "xmp",
2097             "xml"
2098             );
2099
2100     $self_closing_tags =  Array(
2101             "img",
2102             "br",
2103             "hr",
2104             "input",
2105             "outbind"
2106             );
2107
2108     $force_tag_closing = true;
2109
2110     $rm_attnames = Array(
2111             "/.*/" =>
2112             Array(
2113                 "/target/i",
2114                 "/^on.*/i",
2115                 "/^dynsrc/i",
2116                 "/^data.*/i",
2117                 "/^lowsrc.*/i"
2118                 )
2119             );
2120
2121     $secremoveimg = "../images/" . _("sec_remove_eng.png");
2122     $bad_attvals = Array(
2123             "/.*/" =>
2124             Array(
2125                 "/^src|background/i" =>
2126                 Array(
2127                     Array(
2128                         "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2129                         "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2130                         "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2131                         ),
2132                     Array(
2133                         "\\1$secremoveimg\\2",
2134                         "\\1$secremoveimg\\2",
2135                         "\\1$secremoveimg\\2",
2136                         )
2137                     ),
2138                 "/^href|action/i" =>
2139                 Array(
2140                     Array(
2141                         "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2142                         "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2143                         "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2144                         ),
2145                     Array(
2146                         "\\1#\\1",
2147                         "\\1#\\1",
2148                         "\\1#\\1"
2149                         )
2150                     ),
2151         "/^style/i" =>
2152             Array(
2153                 Array(
2154                     "/\/\*.*\*\//",
2155                     "/expression/i",
2156                     "/binding/i",
2157                     "/behaviou*r/i",
2158                     "/include-source/i",
2159                     "/position\s*:\s*absolute/i",
2160                     "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2161                     "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2162                     "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2163                     "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2164                     ),
2165                 Array(
2166                     "",
2167                     "idiocy",
2168                     "idiocy",
2169                     "idiocy",
2170                     "idiocy",
2171                     "",
2172                     "url(\\1#\\1)",
2173                     "url(\\1#\\1)",
2174                     "url(\\1#\\1)",
2175                     "\\1:url(\\2#\\3)"
2176                     )
2177                 )
2178             )
2179         );
2180     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
2181         $view_unsafe_images = false;
2182     }
2183     if (!$view_unsafe_images){
2184         /**
2185          * Remove any references to http/https if view_unsafe_images set
2186          * to false.
2187          */
2188         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2189                 '/^([\'\"])\s*https*:.*([\'\"])/si');
2190         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2191                 "\\1$secremoveimg\\1");
2192         array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2193                 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2194         array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2195                 "url(\\1$secremoveimg\\1)");
2196     }
2197
2198     $add_attr_to_tag = Array(
2199             "/^a$/i" =>
2200             Array('target'=>'"_blank"',
2201                 'title'=>'"'._("This external link will open in a new window").'"'
2202                 )
2203             );
2204     $trusted = sq_sanitize($body,
2205                            $tag_list,
2206                            $rm_tags_with_content,
2207                            $self_closing_tags,
2208                            $force_tag_closing,
2209                            $rm_attnames,
2210                            $bad_attvals,
2211                            $add_attr_to_tag,
2212                            $message,
2213                            $id,
2214                            $mailbox
2215                            );
2216     if (preg_match("|$secremoveimg|i", $trusted)){
2217         $has_unsafe_images = true;
2218     }
2219
2220     // we want to parse mailto's in HTML output, change to SM compose links
2221     // this is a modified version of code from url_parser.php... but Marc is
2222     // right: we need a better filtering implementation; adding this randomly
2223     // here is not a great solution
2224     //
2225     if ($take_mailto_links) {
2226         // parseUrl($trusted);   // this even parses URLs inside of tags... too aggressive
2227         global $MailTo_PReg_Match;
2228         $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1);
2229         if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2230             foreach ($regs[0] as $i => $mailto_before) {
2231                 $mailto_params = $regs[10][$i];
2232                 // get rid of any tailing quote since we have to add send_to to the end
2233                 //
2234                 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2235                     $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2236                 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2237                     $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2238
2239                 if ($regs[1][$i]) {    //if there is an email addr before '?', we need to merge it with the params
2240                     $to = 'to=' . $regs[1][$i];
2241                     if (strpos($mailto_params, 'to=') > -1)    //already a 'to='
2242                         $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2243                     else {
2244                         if ($mailto_params)    //already some params, append to them
2245                             $mailto_params .= '&amp;' . $to;
2246                         else
2247                             $mailto_params .= '?' . $to;
2248                     }
2249                 }
2250
2251                 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2252
2253                 // we'll already have target=_blank, no need to allow comp_in_new
2254                 // here (which would be a lot more work anyway)
2255                 //
2256                 global $compose_new_win;
2257                 $temp_comp_in_new = $compose_new_win;
2258                 $compose_new_win = 0;
2259                 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2260                 $compose_new_win = $temp_comp_in_new;
2261
2262                 // remove <a href=" and anything after the next quote (we only
2263                 // need the uri, not the link HTML) in compose uri
2264                 //
2265                 $comp_uri = substr($comp_uri, 9);
2266                 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2267                 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2268             }
2269         }
2270     }
2271
2272     return $trusted;
2273 }
2274
2275 /**
2276  * function SendDownloadHeaders - send file to the browser
2277  *
2278  * Original Source: SM core src/download.php
2279  * moved here to make it available to other code, and separate
2280  * front end from back end functionality.
2281  *
2282  * @param string $type0 first half of mime type
2283  * @param string $type1 second half of mime type
2284  * @param string $filename filename to tell the browser for downloaded file
2285  * @param boolean $force whether to force the download dialog to pop
2286  * @param optional integer $filesize send the Content-Header and length to the browser
2287  * @return void
2288  */
2289 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2290     global $languages, $squirrelmail_language;
2291     $isIE = $isIE6plus = false;
2292
2293     sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2294
2295     if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2296             strstr($HTTP_USER_AGENT, 'Opera') === false) {
2297         $isIE = true;
2298     }
2299
2300     if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2301         ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2302         $isIE6plus = true;
2303     }
2304
2305     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2306             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2307         $filename =
2308             call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2309     } else {
2310         $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
2311     }
2312
2313     // A Pox on Microsoft and it's Internet Explorer!
2314     //
2315     // IE has lots of bugs with file downloads.
2316     // It also has problems with SSL.  Both of these cause problems
2317     // for us in this function.
2318     //
2319     // See this article on Cache Control headers and SSL
2320     // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2321     //
2322     // The best thing you can do for IE is to upgrade to the latest
2323     // version
2324     //set all the Cache Control Headers for IE
2325     if ($isIE) {
2326         $filename=rawurlencode($filename);
2327         header ("Pragma: public");
2328         header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2329         header ("Cache-Control: post-check=0, pre-check=0", false);
2330         header ("Cache-Control: private");
2331
2332         //set the inline header for IE, we'll add the attachment header later if we need it
2333         header ("Content-Disposition: inline; filename=$filename");
2334     }
2335
2336     if (!$force) {
2337         // Try to show in browser window
2338         header ("Content-Disposition: inline; filename=\"$filename\"");
2339         header ("Content-Type: $type0/$type1; name=\"$filename\"");
2340     } else {
2341         // Try to pop up the "save as" box
2342
2343         // IE makes this hard.  It pops up 2 save boxes, or none.
2344         // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2345         // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2346         // But, according to Microsoft, it is "RFC compliant but doesn't
2347         // take into account some deviations that allowed within the
2348         // specification."  Doesn't that mean RFC non-compliant?
2349         // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2350
2351         // all browsers need the application/octet-stream header for this
2352         header ("Content-Type: application/octet-stream; name=\"$filename\"");
2353
2354         // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2355         // Do not have quotes around filename, but that applied to
2356         // "attachment"... does it apply to inline too?
2357         header ("Content-Disposition: attachment; filename=\"$filename\"");
2358
2359         if ($isIE && !$isIE6plus) {
2360             // This combination seems to work mostly.  IE 5.5 SP 1 has
2361             // known issues (see the Microsoft Knowledge Base)
2362
2363             // This works for most types, but doesn't work with Word files
2364             header ("Content-Type: application/download; name=\"$filename\"");
2365
2366             // These are spares, just in case.  :-)
2367             //header("Content-Type: $type0/$type1; name=\"$filename\"");
2368             //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2369             //header("Content-Type: application/octet-stream; name=\"$filename\"");
2370         } else {
2371             // another application/octet-stream forces download for Netscape
2372             header ("Content-Type: application/octet-stream; name=\"$filename\"");
2373         }
2374     }
2375
2376     //send the content-length header if the calling function provides it
2377     if ($filesize > 0) {
2378         header("Content-Length: $filesize");
2379     }
2380
2381 }  // end fn SendDownloadHeaders
2382
2383 ?>