functions/mime.php

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