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 /**
  16  * dependency information
  17    functions       dependency
  18    mime_structure
  19         class/mime/Message.class.php
  20             Message::parseStructure
  21         functions/page_header.php
  22             displayPageHeader
  23         functions/display_messages.php
  24             plain_error_message
  25    mime_fetch_body
  26         functions/imap_general.php
  27             sqimap_run_command
  28    mime_print_body_lines
  29
  30
  31
  32 functions/imap.php
  33 functions/attachment_common.php
  34 functions/display_messages.php
  35
  36 magicHtml => url_parser
  37 translateText => url_parser
  38
  39 */
  40
  41
  42 /* -------------------------------------------------------------------------- */
  43 /* MIME DECODING                                                              */
  44 /* -------------------------------------------------------------------------- */
  45
  46 /**
  47  * Get the MIME structure
  48  *
  49  * This function gets the structure of a message and stores it in the "message" class.
  50  * It will return this object for use with all relevant header information and
  51  * fully parsed into the standard "message" object format.
  52  */
  53 function mime_structure ($bodystructure, $flags=array()) {
  54
  55     /* Isolate the body structure and remove beginning and end parenthesis. */
  56     $read = trim(substr ($bodystructure, strpos(strtolower($bodystructure), 'bodystructure') + 13));
  57     $read = trim(substr ($read, 0, -1));
  58     $i = 0;
  59     $msg = Message::parseStructure($read,$i);
  60     if (!is_object($msg)) {
  61         global $color, $mailbox;
  62         /* removed urldecode because $_GET is auto urldecoded ??? */
  63         displayPageHeader( $color, $mailbox );
  64         $errormessage  = _("SquirrelMail could not decode the bodystructure of the message");
  65         $errormessage .= '<br />'._("The bodystructure provided by your IMAP server:").'<br /><br />';
  66         $errormessage .= '<pre>' . htmlspecialchars($read) . '</pre>';
  67         plain_error_message( $errormessage, $color );
  68         echo '</body></html>';
  69         exit;
  70     }
  71     if (count($flags)) {
  72         foreach ($flags as $flag) {
  73             $char = strtoupper($flag{1});
  74             switch ($char) {
  75                 case 'S':
  76                     if (strtolower($flag) == '\\seen') {
  77                         $msg->is_seen = true;
  78                     }
  79                     break;
  80                 case 'A':
  81                     if (strtolower($flag) == '\\answered') {
  82                         $msg->is_answered = true;
  83                     }
  84                     break;
  85                 case 'D':
  86                     if (strtolower($flag) == '\\deleted') {
  87                         $msg->is_deleted = true;
  88                     }
  89                     break;
  90                 case 'F':
  91                     if (strtolower($flag) == '\\flagged') {
  92                         $msg->is_flagged = true;
  93                     }
  94                     break;
  95                 case 'M':
  96                     if (strtolower($flag) == '$mdnsent') {
  97                         $msg->is_mdnsent = true;
  98                     }
  99                     break;
 100                 default:
 101                     break;
 102             }
 103         }
 104     }
 105     //    listEntities($msg);
 106     return $msg;
 107 }
 108
 109
 110
 111 /* This starts the parsing of a particular structure.  It is called recursively,
 112  * so it can be passed different structures.  It returns an object of type
 113  * $message.
 114  * First, it checks to see if it is a multipart message.  If it is, then it
 115  * handles that as it sees is necessary.  If it is just a regular entity,
 116  * then it parses it and adds the necessary header information (by calling out
 117  * to mime_get_elements()
 118  */
 119
 120 function mime_fetch_body($imap_stream, $id, $ent_id=1, $fetch_size=0) {
 121     /* Do a bit of error correction.  If we couldn't find the entity id, just guess
 122      * that it is the first one.  That is usually the case anyway.
 123      */
 124
 125     if (!$ent_id) {
 126         $cmd = "FETCH $id BODY[]";
 127     } else {
 128         $cmd = "FETCH $id BODY[$ent_id]";
 129     }
 130
 131     if ($fetch_size!=0) $cmd .= "<0.$fetch_size>";
 132
 133     $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, TRUE);
 134     do {
 135         $topline = trim(array_shift($data));
 136     } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
 137
 138     $wholemessage = implode('', $data);
 139     if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
 140         $ret = substr($wholemessage, 0, $regs[1]);
 141         /* There is some information in the content info header that could be important
 142          * in order to parse html messages. Let's get them here.
 143          */
 144 //        if ($ret{0} == '<') {
 145 //            $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, TRUE);
 146 //        }
 147     } else if (ereg('"([^"]*)"', $topline, $regs)) {
 148         $ret = $regs[1];
 149     } else {
 150         global $where, $what, $mailbox, $passed_id, $startMessage;
 151         $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
 152         if (isset($where) && isset($what)) {
 153             $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
 154         } else {
 155             $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 156         }
 157         $par .= '&amp;response=' . urlencode($response) .
 158             '&amp;message='  . urlencode($message)  .
 159             '&amp;topline='  . urlencode($topline);
 160
 161         echo '<tt><br />' .
 162             '<table width="80%"><tr>' .
 163             '<tr><td colspan="2">' .
 164             _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
 165             '</td></tr>' .
 166             '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
 167             '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
 168             '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
 169             '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
 170             "</table><br /></tt></font><hr />";
 171
 172         $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, TRUE);
 173         array_shift($data);
 174         $wholemessage = implode('', $data);
 175
 176         $ret = $wholemessage;
 177     }
 178     return $ret;
 179 }
 180
 181 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding, $rStream='php://stdout') {
 182
 183     /* Don't kill the connection if the browser is over a dialup
 184      * and it would take over 30 seconds to download it.
 185      * Don't call set_time_limit in safe mode.
 186      */
 187
 188     if (!ini_get('safe_mode')) {
 189         set_time_limit(0);
 190     }
 191     /* in case of base64 encoded attachments, do not buffer them.
 192        Instead, echo the decoded attachment directly to screen */
 193     if (strtolower($encoding) == 'base64') {
 194         if (!$ent_id) {
 195             $query = "FETCH $id BODY[]";
 196         } else {
 197             $query = "FETCH $id BODY[$ent_id]";
 198         }
 199         sqimap_run_command($imap_stream,$query,true,$response,$message,TRUE,'sqimap_base64_decode',$rStream,true);
 200     } else {
 201         $body = mime_fetch_body ($imap_stream, $id, $ent_id);
 202         if (is_resource($rStream)) {
 203             fputs($rStream,decodeBody($body,$encoding));
 204         } else {
 205             echo decodeBody($body, $encoding);
 206         }
 207     }
 208
 209     /*
 210        TODO, use the same method for quoted printable.
 211        However, I assume that quoted printable attachments aren't that large
 212        so the performancegain / memory usage drop will be minimal.
 213        If we decide to add that then we need to adapt sqimap_fread because
 214        we need to split te result on \n and fread doesn't stop at \n. That
 215        means we also should provide $results from sqimap_fread (by ref) to
 216        te function and set $no_return to false. The $filter function for
 217        quoted printable should handle unsetting of $results.
 218      */
 219     /*
 220        TODO 2: find out how we write to the output stream php://stdout. fwrite
 221        doesn't work because 'php://stdout isn't a stream.
 222      */
 223
 224     return;
 225 }
 226
 227 /* -[ END MIME DECODING ]----------------------------------------------------------- */
 228
 229 /* This is here for debugging purposes.  It will print out a list
 230  * of all the entity IDs that are in the $message object.
 231  */
 232 function listEntities ($message) {
 233     if ($message) {
 234         echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br />';
 235         for ($i = 0; isset($message->entities[$i]); $i++) {
 236             echo "$i : ";
 237             $msg = listEntities($message->entities[$i]);
 238
 239             if ($msg) {
 240                 echo "return: ";
 241                 return $msg;
 242             }
 243         }
 244     }
 245 }
 246
 247 function getPriorityStr($priority) {
 248     $priority_level = substr($priority,0,1);
 249
 250     switch($priority_level) {
 251         /* Check for a higher then normal priority. */
 252         case '1':
 253         case '2':
 254             $priority_string = _("High");
 255             break;
 256
 257         /* Check for a lower then normal priority. */
 258         case '4':
 259         case '5':
 260             $priority_string = _("Low");
 261             break;
 262
 263         /* Check for a normal priority. */
 264         case '3':
 265         default:
 266             $priority_level = '3';
 267             $priority_string = _("Normal");
 268             break;
 269
 270     }
 271     return $priority_string;
 272 }
 273
 274 /* returns a $message object for a particular entity id */
 275 function getEntity ($message, $ent_id) {
 276     return $message->getEntity($ent_id);
 277 }
 278
 279 /* translateText
 280  * Extracted from strings.php 23/03/2002
 281  */
 282
 283 function translateText(&$body, $wrap_at, $charset) {
 284     global $where, $what;   /* from searching */
 285     global $color;          /* color theme */
 286
 287     // require_once(SM_PATH . 'functions/url_parser.php');
 288
 289     $body_ary = explode("\n", $body);
 290     for ($i=0; $i < count($body_ary); $i++) {
 291         $line = $body_ary[$i];
 292         if (strlen($line) - 2 >= $wrap_at) {
 293             sqWordWrap($line, $wrap_at, $charset);
 294         }
 295         $line = charset_decode($charset, $line);
 296         $line = str_replace("\t", '        ', $line);
 297
 298         parseUrl ($line);
 299
 300         $quotes = 0;
 301         $pos = 0;
 302         $j = strlen($line);
 303
 304         while ($pos < $j) {
 305             if ($line[$pos] == ' ') {
 306                 $pos++;
 307             } else if (strpos($line, '&gt;', $pos) === $pos) {
 308                 $pos += 4;
 309                 $quotes++;
 310             } else {
 311                 break;
 312             }
 313         }
 314
 315         if ($quotes % 2) {
 316             $line = '<span class="quote1">' . $line . '</span>';
 317         } elseif ($quotes) {
 318             $line = '<span class="quote2">' . $line . '</span>';
 319         }
 320
 321         $body_ary[$i] = $line;
 322     }
 323     $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
 324 }
 325
 326 /**
 327  * This returns a parsed string called $body. That string can then
 328  * be displayed as the actual message in the HTML. It contains
 329  * everything needed, including HTML Tags, Attachments at the
 330  * bottom, etc.
 331  *
 332  * Since 1.2.0 function uses message_body hook.
 333  * Till 1.3.0 function included output of formatAttachments().
 334  *
 335  * @param resource $imap_stream imap connection resource
 336  * @param object $message squirrelmail message object
 337  * @param array $color squirrelmail color theme array
 338  * @param integer $wrap_at number of characters per line
 339  * @param string $ent_num (since 1.3.0) message part id
 340  * @param integer $id (since 1.3.0) message id
 341  * @param string $mailbox (since 1.3.0) imap folder name
 342  * @param boolean $clean (since 1.5.1) Do not output stuff that's irrelevant for the printable version.
 343  * @return string html formated message text
 344  */
 345 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX', $clean=FALSE) {
 346     /* This if statement checks for the entity to show as the
 347      * primary message. To add more of them, just put them in the
 348      * order that is their priority.
 349      */
 350     global $startMessage, $languages, $squirrelmail_language,
 351            $show_html_default, $sort, $has_unsafe_images, $passed_ent_id,
 352            $use_iframe, $iframe_height, $download_and_unsafe_link,
 353            $download_href, $unsafe_image_toggle_href, $unsafe_image_toggle_text;
 354
 355     // workaround for not updated config.php
 356     if (! isset($use_iframe)) $use_iframe = false;
 357
 358     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
 359         $view_unsafe_images = false;
 360     }
 361
 362     $body = '';
 363     $urlmailbox = urlencode($mailbox);
 364     $body_message = getEntity($message, $ent_num);
 365     if (($body_message->header->type0 == 'text') ||
 366             ($body_message->header->type0 == 'rfc822')) {
 367         $body = mime_fetch_body ($imap_stream, $id, $ent_num);
 368         $body = decodeBody($body, $body_message->header->encoding);
 369
 370         if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 371                 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
 372             if (mb_detect_encoding($body) != 'ASCII') {
 373                 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
 374             }
 375         }
 376         $hookResults = do_hook("message_body", $body);
 377         $body = $hookResults[1];
 378
 379         /* If there are other types that shouldn't be formatted, add
 380          * them here.
 381          */
 382
 383         if ($body_message->header->type1 == 'html') {
 384             if ($show_html_default <> 1) {
 385                 $entity_conv = array('&nbsp;' => ' ',
 386                                      '<p>'    => "\n",
 387                                      '<P>'    => "\n",
 388                                      '<br>'   => "\n",
 389                                      '<BR>'   => "\n",
 390                                      '<br />' => "\n",
 391                                      '<BR />' => "\n",
 392                                      '&gt;'   => '>',
 393                                      '&lt;'   => '<');
 394                 $body = strtr($body, $entity_conv);
 395                 $body = strip_tags($body);
 396                 $body = trim($body);
 397                 translateText($body, $wrap_at,
 398                         $body_message->header->getParameter('charset'));
 399             } elseif ($use_iframe && ! $clean) {
 400                 // $clean is used to remove iframe in printable view.
 401
 402                 /**
 403                  * If we don't add html message between iframe tags,
 404                  * we must detect unsafe images and modify $has_unsafe_images.
 405                  */
 406                 $html_body = magicHTML($body, $id, $message, $mailbox);
 407                 // Convert character set in order to display html mails in different character set
 408                 $html_body = charset_decode($body_message->header->getParameter('charset'),$html_body,false,true);
 409
 410                 // creating iframe url
 411                 $iframeurl=sqm_baseuri().'src/view_html.php?'
 412                     . 'mailbox=' . $urlmailbox
 413                     . '&amp;passed_id=' . $id
 414                     . '&amp;ent_id=' . $ent_num
 415                     . '&amp;view_unsafe_images=' . (int) $view_unsafe_images;
 416
 417                 global $oTemplate;
 418                 $oTemplate->assign('iframe_url', $iframeurl);
 419                 $oTemplate->assign('html_body', $html_body);
 420
 421                 $body = $oTemplate->fetch('read_html_iframe.tpl');
 422             } else {
 423                 // old way of html rendering
 424                 $body = magicHTML($body, $id, $message, $mailbox);
 425                 /**
 426                  * convert character set. charset_decode does not remove html special chars
 427                  * applied by magicHTML functions and does not sanitize them second time if
 428                  * fourth argument is true.
 429                  */
 430                 $body = charset_decode($body_message->header->getParameter('charset'),$body,false,true);
 431             }
 432         } else {
 433             translateText($body, $wrap_at,
 434                     $body_message->header->getParameter('charset'));
 435         }
 436
 437         // if this is the clean display (i.e. printer friendly), stop here.
 438         if ( $clean ) {
 439             return $body;
 440         }
 441
 442         $download_and_unsafe_link = '';
 443
 444         $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
 445             '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
 446             '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 447         if (isset($passed_ent_id)) {
 448             $link .= '&amp;passed_ent_id='.$passed_ent_id;
 449         }
 450         $download_href = SM_PATH . 'src/download.php?absolute_dl=true&amp;' . $link;
 451         $download_and_unsafe_link .= '&nbsp;|&nbsp;<a href="'. $download_href .'">' . _("Download this as a file") .  '</a>';
 452         if ($view_unsafe_images) {
 453             $text = _("Hide Unsafe Images");
 454         } else {
 455             if (isset($has_unsafe_images) && $has_unsafe_images) {
 456                 $link .= '&amp;view_unsafe_images=1';
 457                 $text = _("View Unsafe Images");
 458             } else {
 459                 $text = '';
 460             }
 461         }
 462         if($text != '') {
 463             $unsafe_image_toggle_href = SM_PATH . 'src/read_body.php?'.$link;
 464             $unsafe_image_toggle_text = $text;
 465             $download_and_unsafe_link .= '&nbsp;|&nbsp;<a href="'. $unsafe_image_toggle_href .'">' . $text . '</a>';
 466         }
 467     }
 468     return $body;
 469 }
 470
 471 /**
 472  * Generate attachments array for passing to templates.  Separated from
 473  * formatAttachments() below so that the same array can be given to the
 474  * print-friendly version.
 475  *
 476  * @since 1.5.2
 477  * @param object $message SquirrelMail message object
 478  * @param array $exclude_id message parts that are not attachments.
 479  * @param string $mailbox mailbox name
 480  * @param integer $id message id
 481  */
 482 function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
 483     global $where, $what, $startMessage, $color, $passed_ent_id, $base_uri;
 484
 485     $att_ar = $message->getAttachments($exclude_id);
 486     $urlMailbox = urlencode($mailbox);
 487
 488     $attachments = array();
 489     foreach ($att_ar as $att) {
 490         $ent = $att->entity_id;
 491         $header = $att->header;
 492         $type0 = strtolower($header->type0);
 493         $type1 = strtolower($header->type1);
 494         $name = '';
 495         $links = array();
 496         $links['download link']['text'] = _("Download");
 497         $links['download link']['href'] = $base_uri .
 498             "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 499
 500         if ($type0 =='message' && $type1 == 'rfc822') {
 501             $default_page = $base_uri  . 'src/read_body.php';
 502             $rfc822_header = $att->rfc822_header;
 503             $filename = $rfc822_header->subject;
 504             if (trim( $filename ) == '') {
 505                 $filename = 'untitled-[' . $ent . ']' ;
 506             }
 507             $from_o = $rfc822_header->from;
 508             if (is_object($from_o)) {
 509                 $from_name = decodeHeader($from_o->getAddress(false));
 510             } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
 511                 // something weird happens when a digest message is opened and you return to the digest
 512                 // now the from object is part of an array. Probably the parseHeader call overwrites the info
 513                 // retrieved from the bodystructure in a different way. We need to fix this later.
 514                 // possible starting point, do not fetch header we already have and inspect how
 515                 // the rfc822_header object behaves.
 516                 $from_name = decodeHeader($from_o[0]->getAddress(false));
 517             } else {
 518                 $from_name = _("Unknown sender");
 519             }
 520             $description = _("From").': '.$from_name;
 521         } else {
 522             $default_page = $base_uri  . 'src/download.php';
 523             $filename = $att->getFilename();
 524             if ($header->description) {
 525                 $description = decodeHeader($header->description);
 526             } else {
 527                 $description = '';
 528             }
 529         }
 530
 531         $display_filename = $filename;
 532         if (isset($passed_ent_id)) {
 533             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 534         } else {
 535             $passed_ent_id_link = '';
 536         }
 537         $defaultlink = $default_page . "?startMessage=$startMessage"
 538             . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 539             . '&amp;ent_id='.$ent.$passed_ent_id_link;
 540         if ($where && $what) {
 541             $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 542         }
 543
 544         /* This executes the attachment hook with a specific MIME-type.
 545          * If that doesn't have results, it tries if there's a rule
 546          * for a more generic type.
 547          */
 548         $hookresults = do_hook("attachment $type0/$type1", $links,
 549                 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 550                 $display_filename, $where, $what);
 551         if(count($hookresults[1]) <= 1) {
 552             $hookresults = do_hook("attachment $type0/*", $links,
 553                     $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 554                     $display_filename, $where, $what);
 555         }
 556
 557         $links = $hookresults[1];
 558         $defaultlink = $hookresults[6];
 559
 560         $this_attachment = array();
 561         $this_attachment['Name'] = decodeHeader($display_filename);
 562         $this_attachment['Description'] = $description;
 563         $this_attachment['DefaultHREF'] = $defaultlink;
 564         $this_attachment['DownloadHREF'] = $links['download link']['href'];
 565         $this_attachment['ViewHREF'] = isset($links['attachment_common']) ? $links['attachment_common']['href'] : '';
 566         $this_attachment['Size'] = $header->size;
 567         $this_attachment['ContentType'] = htmlspecialchars($type0 .'/'. $type1);
 568         $this_attachment['OtherLinks'] = array();
 569         foreach ($links as $val) {
 570             if ($val['text']==_("Download") || $val['text'] == _("View"))
 571                 continue;
 572             if (empty($val['text']) && empty($val['extra']))
 573                 continue;
 574
 575             $temp = array();
 576             $temp['HREF'] = $val['href'];
 577             $temp['Text'] = (empty($val['text']) ? '' : $val['text']) . (empty($val['extra']) ? '' : $val['extra']);
 578             $this_attachment['OtherLinks'][] = $temp;
 579         }
 580         $attachments[] = $this_attachment;
 581
 582         unset($links);
 583     }
 584
 585     return $attachments;
 586 }
 587
 588 /**
 589  * Displays attachment links and information
 590  *
 591  * Since 1.3.0 function is not included in formatBody() call.
 592  *
 593  * Since 1.0.2 uses attachment $type0/$type1 hook.
 594  * Since 1.2.5 uses attachment $type0/* hook.
 595  * Since 1.5.0 uses attachments_bottom hook.
 596  * Since 1.5.2 uses templates and does *not* return a value.
 597  *
 598  * @param object $message SquirrelMail message object
 599  * @param array $exclude_id message parts that are not attachments.
 600  * @param string $mailbox mailbox name
 601  * @param integer $id message id
 602  */
 603 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 604     global $oTemplate;
 605
 606     $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
 607
 608     $oTemplate->assign('attachments', $attach);
 609     $oTemplate->display('read_attachments.tpl');
 610 }
 611
 612 function sqimap_base64_decode(&$string) {
 613
 614     // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
 615     // fly decoding (to reduce memory usage) you have to check if the
 616     // data has incomplete pairs
 617
 618     // Remove the noise in order to check if the 4 bytes pairs are complete
 619     $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
 620
 621     $sStringRem = '';
 622     $iMod = strlen($string) % 4;
 623     if ($iMod) {
 624         $sStringRem = substr($string,-$iMod);
 625         // Check if $sStringRem contains padding characters
 626         if (substr($sStringRem,-1) != '=') {
 627             $string = substr($string,0,-$iMod);
 628         } else {
 629             $sStringRem = '';
 630         }
 631     }
 632     $string = base64_decode($string);
 633     return $sStringRem;
 634 }
 635
 636 /**
 637  * Decodes encoded message body
 638  *
 639  * This function decodes the body depending on the encoding type.
 640  * Currently quoted-printable and base64 encodings are supported.
 641  * decode_body hook was added to this function in 1.4.2/1.5.0
 642  * @param string $body encoded message body
 643  * @param string $encoding used encoding
 644  * @return string decoded string
 645  * @since 1.0
 646  */
 647 function decodeBody($body, $encoding) {
 648
 649     $body = str_replace("\r\n", "\n", $body);
 650     $encoding = strtolower($encoding);
 651
 652     $encoding_handler = do_hook_function('decode_body', $encoding);
 653
 654
 655     // plugins get first shot at decoding the body
 656     //
 657     if (!empty($encoding_handler) && function_exists($encoding_handler)) {
 658         $body = $encoding_handler('decode', $body);
 659
 660     } elseif ($encoding == 'quoted-printable' ||
 661             $encoding == 'quoted_printable') {
 662         /**
 663          * quoted_printable_decode() function is broken in older
 664          * php versions. Text with \r\n decoding was fixed only
 665          * in php 4.3.0. Minimal code requirement 4.0.4 +
 666          * str_replace("\r\n", "\n", $body); call.
 667          */
 668         $body = quoted_printable_decode($body);
 669     } elseif ($encoding == 'base64') {
 670         $body = base64_decode($body);
 671     }
 672
 673     // All other encodings are returned raw.
 674     return $body;
 675 }
 676
 677 /**
 678  * Decodes headers
 679  *
 680  * This functions decode strings that is encoded according to
 681  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 682  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 683  *
 684  * @param string $string header string that has to be made readable
 685  * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
 686  * @param boolean $htmlsave preserve spaces and sanitize html special characters. defaults to true
 687  * @param boolean $decide decide if string can be utfencoded. defaults to false
 688  * @return string decoded header string
 689  */
 690 function decodeHeader ($string, $utfencode=true,$htmlsave=true,$decide=false) {
 691     global $languages, $squirrelmail_language,$default_charset;
 692     if (is_array($string)) {
 693         $string = implode("\n", $string);
 694     }
 695
 696     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 697             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
 698         $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
 699         // Do we need to return at this point?
 700         // return $string;
 701     }
 702     $i = 0;
 703     $iLastMatch = -2;
 704     $encoded = true;
 705
 706     $aString = explode(' ',$string);
 707     $ret = '';
 708     foreach ($aString as $chunk) {
 709         if ($encoded && $chunk === '') {
 710             continue;
 711         } elseif ($chunk === '') {
 712             $ret .= ' ';
 713             continue;
 714         }
 715         $encoded = false;
 716         /* if encoded words are not separated by a linear-space-white we still catch them */
 717         $j = $i-1;
 718
 719         while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 720             /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
 721             if ($iLastMatch !== $j) {
 722                 if ($htmlsave) {
 723                     $ret .= '&#32;';
 724                 } else {
 725                     $ret .= ' ';
 726                 }
 727             }
 728             $iLastMatch = $i;
 729             $j = $i;
 730             if ($htmlsave) {
 731                 $ret .= htmlspecialchars($res[1]);
 732             } else {
 733                 $ret .= $res[1];
 734             }
 735             $encoding = ucfirst($res[3]);
 736
 737             /* decide about valid decoding */
 738             if ($decide && is_conversion_safe($res[2])) {
 739                 $utfencode=true;
 740                 $can_be_encoded=true;
 741             } else {
 742                 $can_be_encoded=false;
 743             }
 744             switch ($encoding)
 745             {
 746                 case 'B':
 747                     $replace = base64_decode($res[4]);
 748                     if ($utfencode) {
 749                         if ($can_be_encoded) {
 750                             /* convert string to different charset,
 751                              * if functions asks for it (usually in compose)
 752                              */
 753                             $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsave);
 754                         } else {
 755                             // convert string to html codes in order to display it
 756                             $ret .= charset_decode($res[2],$replace);
 757                         }
 758                     } else {
 759                         if ($htmlsave) {
 760                             $replace = htmlspecialchars($replace);
 761                         }
 762                         $ret.= $replace;
 763                     }
 764                     break;
 765                 case 'Q':
 766                     $replace = str_replace('_', ' ', $res[4]);
 767                     $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 768                             $replace);
 769                     if ($utfencode) {
 770                         if ($can_be_encoded) {
 771                             /* convert string to different charset,
 772                              * if functions asks for it (usually in compose)
 773                              */
 774                             $replace = charset_convert($res[2], $replace,$default_charset,$htmlsave);
 775                         } else {
 776                             // convert string to html codes in order to display it
 777                             $replace = charset_decode($res[2], $replace);
 778                         }
 779                     } else {
 780                         if ($htmlsave) {
 781                             $replace = htmlspecialchars($replace);
 782                         }
 783                     }
 784                     $ret .= $replace;
 785                     break;
 786                 default:
 787                     break;
 788             }
 789             $chunk = $res[5];
 790             $encoded = true;
 791         }
 792         if (!$encoded) {
 793             if ($htmlsave) {
 794                 $ret .= '&#32;';
 795             } else {
 796                 $ret .= ' ';
 797             }
 798         }
 799
 800         if (!$encoded && $htmlsave) {
 801             $ret .= htmlspecialchars($chunk);
 802         } else {
 803             $ret .= $chunk;
 804         }
 805         ++$i;
 806     }
 807     /* remove the first added space */
 808     if ($ret) {
 809         if ($htmlsave) {
 810             $ret = substr($ret,5);
 811         } else {
 812             $ret = substr($ret,1);
 813         }
 814     }
 815
 816     return $ret;
 817 }
 818
 819 /**
 820  * Encodes header
 821  *
 822  * Function uses XTRA_CODE _encodeheader function, if such function exists.
 823  *
 824  * Function uses Q encoding by default and encodes a string according to RFC
 825  * 1522 for use in headers if it contains 8-bit characters or anything that
 826  * looks like it should be encoded.
 827  *
 828  * Function switches to B encoding and encodeHeaderBase64() function, if
 829  * string is 8bit and multibyte character set supported by mbstring extension
 830  * is used. It can cause E_USER_NOTICE errors, if interface is used with
 831  * multibyte character set unsupported by mbstring extension.
 832  *
 833  * @param string $string header string, that has to be encoded
 834  * @return string quoted-printable encoded string
 835  * @todo make $mb_charsets system wide constant
 836  */
 837 function encodeHeader ($string) {
 838     global $default_charset, $languages, $squirrelmail_language;
 839
 840     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 841             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
 842         return  call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
 843     }
 844
 845     // Use B encoding for multibyte charsets
 846     $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
 847     if (in_array($default_charset,$mb_charsets) &&
 848         in_array($default_charset,sq_mb_list_encodings()) &&
 849         sq_is8bit($string)) {
 850         return encodeHeaderBase64($string,$default_charset);
 851     } elseif (in_array($default_charset,$mb_charsets) &&
 852               sq_is8bit($string) &&
 853               ! in_array($default_charset,sq_mb_list_encodings())) {
 854         // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
 855         // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
 856     }
 857
 858     // Encode only if the string contains 8-bit characters or =?
 859     $j = strlen($string);
 860     $max_l = 75 - strlen($default_charset) - 7;
 861     $aRet = array();
 862     $ret = '';
 863     $iEncStart = $enc_init = false;
 864     $cur_l = $iOffset = 0;
 865     for($i = 0; $i < $j; ++$i) {
 866         switch($string{$i})
 867         {
 868             case '=':
 869             case '<':
 870             case '>':
 871             case ',':
 872             case '?':
 873             case '_':
 874                 if ($iEncStart === false) {
 875                     $iEncStart = $i;
 876                 }
 877                 $cur_l+=3;
 878                 if ($cur_l > ($max_l-2)) {
 879                     /* if there is an stringpart that doesn't need encoding, add it */
 880                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 881                     $aRet[] = "=?$default_charset?Q?$ret?=";
 882                     $iOffset = $i;
 883                     $cur_l = 0;
 884                     $ret = '';
 885                     $iEncStart = false;
 886                 } else {
 887                     $ret .= sprintf("=%02X",ord($string{$i}));
 888                 }
 889                 break;
 890             case '(':
 891             case ')':
 892                 if ($iEncStart !== false) {
 893                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 894                     $aRet[] = "=?$default_charset?Q?$ret?=";
 895                     $iOffset = $i;
 896                     $cur_l = 0;
 897                     $ret = '';
 898                     $iEncStart = false;
 899                 }
 900                 break;
 901             case ' ':
 902                 if ($iEncStart !== false) {
 903                     $cur_l++;
 904                     if ($cur_l > $max_l) {
 905                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 906                         $aRet[] = "=?$default_charset?Q?$ret?=";
 907                         $iOffset = $i;
 908                         $cur_l = 0;
 909                         $ret = '';
 910                         $iEncStart = false;
 911                     } else {
 912                         $ret .= '_';
 913                     }
 914                 }
 915                 break;
 916             default:
 917                 $k = ord($string{$i});
 918                 if ($k > 126) {
 919                     if ($iEncStart === false) {
 920                         // do not start encoding in the middle of a string, also take the rest of the word.
 921                         $sLeadString = substr($string,0,$i);
 922                         $aLeadString = explode(' ',$sLeadString);
 923                         $sToBeEncoded = array_pop($aLeadString);
 924                         $iEncStart = $i - strlen($sToBeEncoded);
 925                         $ret .= $sToBeEncoded;
 926                         $cur_l += strlen($sToBeEncoded);
 927                     }
 928                     $cur_l += 3;
 929                     /* first we add the encoded string that reached it's max size */
 930                     if ($cur_l > ($max_l-2)) {
 931                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 932                         $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
 933                         $cur_l = 3;
 934                         $ret = '';
 935                         $iOffset = $i;
 936                         $iEncStart = $i;
 937                     }
 938                     $enc_init = true;
 939                     $ret .= sprintf("=%02X", $k);
 940                 } else {
 941                     if ($iEncStart !== false) {
 942                         $cur_l++;
 943                         if ($cur_l > $max_l) {
 944                             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 945                             $aRet[] = "=?$default_charset?Q?$ret?=";
 946                             $iEncStart = false;
 947                             $iOffset = $i;
 948                             $cur_l = 0;
 949                             $ret = '';
 950                         } else {
 951                             $ret .= $string{$i};
 952                         }
 953                     }
 954                 }
 955                 break;
 956         }
 957     }
 958
 959     if ($enc_init) {
 960         if ($iEncStart !== false) {
 961             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 962             $aRet[] = "=?$default_charset?Q?$ret?=";
 963         } else {
 964             $aRet[] = substr($string,$iOffset);
 965         }
 966         $string = implode('',$aRet);
 967     }
 968     return $string;
 969 }
 970
 971 /**
 972  * Encodes string according to rfc2047 B encoding header formating rules
 973  *
 974  * It is recommended way to encode headers with character sets that store
 975  * symbols in more than one byte.
 976  *
 977  * Function requires mbstring support. If required mbstring functions are missing,
 978  * function returns false and sets E_USER_WARNING level error message.
 979  *
 980  * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
 981  * that mbstring functions will generate E_WARNING errors, if unsupported
 982  * character set is used. mb_encode_mimeheader function provided by php
 983  * mbstring extension is not used in order to get better control of header
 984  * encoding.
 985  *
 986  * Used php code functions - function_exists(), trigger_error(), strlen()
 987  * (is used with charset names and base64 strings). Used php mbstring
 988  * functions - mb_strlen and mb_substr.
 989  *
 990  * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
 991  * encoding), rfc 2822 (header folding)
 992  *
 993  * @param string $string header string that must be encoded
 994  * @param string $charset character set. Must be supported by mbstring extension.
 995  * Use sq_mb_list_encodings() to detect supported charsets.
 996  * @return string string encoded according to rfc2047 B encoding formating rules
 997  * @since 1.5.1
 998  * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
 999  * @todo Do we want to control max length of header?
1000  * @todo Do we want to control EOL (end-of-line) marker?
1001  * @todo Do we want to translate error message?
1002  */
1003 function encodeHeaderBase64($string,$charset) {
1004     /**
1005      * Check mbstring function requirements.
1006      */
1007     if (! function_exists('mb_strlen') ||
1008         ! function_exists('mb_substr')) {
1009         // set E_USER_WARNING
1010         trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
1011         // return false
1012         return false;
1013     }
1014
1015     // initial return array
1016     $aRet = array();
1017
1018     /**
1019      * header length = 75 symbols max (same as in encodeHeader)
1020      * remove $charset length
1021      * remove =? ? ?= (5 chars)
1022      * remove 2 more chars (\r\n ?)
1023      */
1024     $iMaxLength = 75 - strlen($charset) - 7;
1025
1026     // set first character position
1027     $iStartCharNum = 0;
1028
1029     // loop through all characters. count characters and not bytes.
1030     for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1031         // encode string from starting character to current character.
1032         $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1033
1034         // Check encoded string length
1035         if(strlen($encoded_string)>$iMaxLength) {
1036             // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1037             $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1038
1039             // set new starting character
1040             $iStartCharNum = $iCharNum-1;
1041
1042             // encode last char (in case it is last character in string)
1043             $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1044         } // if string is shorter than max length - add next character
1045     }
1046
1047     // add last encoded string to array
1048     $aRet[] = $encoded_string;
1049
1050     // set initial return string
1051     $sRet = '';
1052
1053     // loop through encoded strings
1054     foreach($aRet as $string) {
1055         // TODO: Do we want to control EOL (end-of-line) marker
1056         if ($sRet!='') $sRet.= " ";
1057
1058         // add header tags and encoded string to return string
1059         $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1060     }
1061
1062     return $sRet;
1063 }
1064
1065 /* This function trys to locate the entity_id of a specific mime element */
1066 function find_ent_id($id, $message) {
1067     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1068         if ($message->entities[$i]->header->type0 == 'multipart')  {
1069             $ret = find_ent_id($id, $message->entities[$i]);
1070         } else {
1071             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
1072 //                if (sq_check_save_extension($message->entities[$i])) {
1073                 return $message->entities[$i]->entity_id;
1074 //                }
1075             } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1076                 /**
1077                  * This is part of a fix for Outlook Express 6.x generating
1078                  * cid URLs without creating content-id headers
1079                  * @@JA - 20050207
1080                  */
1081                 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1082                     return $message->entities[$i]->entity_id;
1083                 }
1084             }
1085         }
1086     }
1087     return $ret;
1088 }
1089
1090 function sq_check_save_extension($message) {
1091     $filename = $message->getFilename();
1092     $ext = substr($filename, strrpos($filename,'.')+1);
1093     $save_extensions = array('jpg','jpeg','gif','png','bmp');
1094     return in_array($ext, $save_extensions);
1095 }
1096
1097
1098 /**
1099  ** HTMLFILTER ROUTINES
1100  */
1101
1102 /**
1103  * This function checks attribute values for entity-encoded values
1104  * and returns them translated into 8-bit strings so we can run
1105  * checks on them.
1106  *
1107  * @param  $attvalue A string to run entity check against.
1108  * @return           Nothing, modifies a reference value.
1109  */
1110 function sq_defang(&$attvalue){
1111     $me = 'sq_defang';
1112     /**
1113      * Skip this if there aren't ampersands or backslashes.
1114      */
1115     if (strpos($attvalue, '&') === false
1116         && strpos($attvalue, '\\') === false){
1117         return;
1118     }
1119     $m = false;
1120     do {
1121         $m = false;
1122         $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1123         $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1124         $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1125     } while ($m == true);
1126     $attvalue = stripslashes($attvalue);
1127 }
1128
1129 /**
1130  * Kill any tabs, newlines, or carriage returns. Our friends the
1131  * makers of the browser with 95% market value decided that it'd
1132  * be funny to make "java[tab]script" be just as good as "javascript".
1133  *
1134  * @param  attvalue  The attribute value before extraneous spaces removed.
1135  * @return attvalue  Nothing, modifies a reference value.
1136  */
1137 function sq_unspace(&$attvalue){
1138     $me = 'sq_unspace';
1139     if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1140         $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1141                                 Array('',   '',   '',   '',   ''), $attvalue);
1142     }
1143 }
1144
1145 /**
1146  * This function returns the final tag out of the tag name, an array
1147  * of attributes, and the type of the tag. This function is called by
1148  * sq_sanitize internally.
1149  *
1150  * @param  $tagname  the name of the tag.
1151  * @param  $attary   the array of attributes and their values
1152  * @param  $tagtype  The type of the tag (see in comments).
1153  * @return           a string with the final tag representation.
1154  */
1155 function sq_tagprint($tagname, $attary, $tagtype){
1156     $me = 'sq_tagprint';
1157
1158     if ($tagtype == 2){
1159         $fulltag = '</' . $tagname . '>';
1160     } else {
1161         $fulltag = '<' . $tagname;
1162         if (is_array($attary) && sizeof($attary)){
1163             $atts = Array();
1164             while (list($attname, $attvalue) = each($attary)){
1165                 array_push($atts, "$attname=$attvalue");
1166             }
1167             $fulltag .= ' ' . join(" ", $atts);
1168         }
1169         if ($tagtype == 3){
1170             $fulltag .= ' /';
1171         }
1172         $fulltag .= '>';
1173     }
1174     return $fulltag;
1175 }
1176
1177 /**
1178  * A small helper function to use with array_walk. Modifies a by-ref
1179  * value and makes it lowercase.
1180  *
1181  * @param  $val a value passed by-ref.
1182  * @return      void since it modifies a by-ref value.
1183  */
1184 function sq_casenormalize(&$val){
1185     $val = strtolower($val);
1186 }
1187
1188 /**
1189  * This function skips any whitespace from the current position within
1190  * a string and to the next non-whitespace value.
1191  *
1192  * @param  $body   the string
1193  * @param  $offset the offset within the string where we should start
1194  *                 looking for the next non-whitespace character.
1195  * @return         the location within the $body where the next
1196  *                 non-whitespace char is located.
1197  */
1198 function sq_skipspace($body, $offset){
1199     $me = 'sq_skipspace';
1200     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1201     if (sizeof($matches{1})){
1202         $count = strlen($matches{1});
1203         $offset += $count;
1204     }
1205     return $offset;
1206 }
1207
1208 /**
1209  * This function looks for the next character within a string.  It's
1210  * really just a glorified "strpos", except it catches if failures
1211  * nicely.
1212  *
1213  * @param  $body   The string to look for needle in.
1214  * @param  $offset Start looking from this position.
1215  * @param  $needle The character/string to look for.
1216  * @return         location of the next occurance of the needle, or
1217  *                 strlen($body) if needle wasn't found.
1218  */
1219 function sq_findnxstr($body, $offset, $needle){
1220     $me  = 'sq_findnxstr';
1221     $pos = strpos($body, $needle, $offset);
1222     if ($pos === FALSE){
1223         $pos = strlen($body);
1224     }
1225     return $pos;
1226 }
1227
1228 /**
1229  * This function takes a PCRE-style regexp and tries to match it
1230  * within the string.
1231  *
1232  * @param  $body   The string to look for needle in.
1233  * @param  $offset Start looking from here.
1234  * @param  $reg    A PCRE-style regex to match.
1235  * @return         Returns a false if no matches found, or an array
1236  *                 with the following members:
1237  *                 - integer with the location of the match within $body
1238  *                 - string with whatever content between offset and the match
1239  *                 - string with whatever it is we matched
1240  */
1241 function sq_findnxreg($body, $offset, $reg){
1242     $me = 'sq_findnxreg';
1243     $matches = Array();
1244     $retarr = Array();
1245     preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1246     if (!isset($matches{0}) || !$matches{0}){
1247         $retarr = false;
1248     } else {
1249         $retarr{0} = $offset + strlen($matches{1});
1250         $retarr{1} = $matches{1};
1251         $retarr{2} = $matches{2};
1252     }
1253     return $retarr;
1254 }
1255
1256 /**
1257  * This function looks for the next tag.
1258  *
1259  * @param  $body   String where to look for the next tag.
1260  * @param  $offset Start looking from here.
1261  * @return         false if no more tags exist in the body, or
1262  *                 an array with the following members:
1263  *                 - string with the name of the tag
1264  *                 - array with attributes and their values
1265  *                 - integer with tag type (1, 2, or 3)
1266  *                 - integer where the tag starts (starting "<")
1267  *                 - integer where the tag ends (ending ">")
1268  *                 first three members will be false, if the tag is invalid.
1269  */
1270 function sq_getnxtag($body, $offset){
1271     $me = 'sq_getnxtag';
1272     if ($offset > strlen($body)){
1273         return false;
1274     }
1275     $lt = sq_findnxstr($body, $offset, "<");
1276     if ($lt == strlen($body)){
1277         return false;
1278     }
1279     /**
1280      * We are here:
1281      * blah blah <tag attribute="value">
1282      * \---------^
1283      */
1284     $pos = sq_skipspace($body, $lt+1);
1285     if ($pos >= strlen($body)){
1286         return Array(false, false, false, $lt, strlen($body));
1287     }
1288     /**
1289      * There are 3 kinds of tags:
1290      * 1. Opening tag, e.g.:
1291      *    <a href="blah">
1292      * 2. Closing tag, e.g.:
1293      *    </a>
1294      * 3. XHTML-style content-less tag, e.g.:
1295      *    <img src="blah" />
1296      */
1297     $tagtype = false;
1298     switch (substr($body, $pos, 1)){
1299         case '/':
1300             $tagtype = 2;
1301             $pos++;
1302             break;
1303         case '!':
1304             /**
1305              * A comment or an SGML declaration.
1306              */
1307             if (substr($body, $pos+1, 2) == "--"){
1308                 $gt = strpos($body, "-->", $pos);
1309                 if ($gt === false){
1310                     $gt = strlen($body);
1311                 } else {
1312                     $gt += 2;
1313                 }
1314                 return Array(false, false, false, $lt, $gt);
1315             } else {
1316                 $gt = sq_findnxstr($body, $pos, ">");
1317                 return Array(false, false, false, $lt, $gt);
1318             }
1319             break;
1320         default:
1321             /**
1322              * Assume tagtype 1 for now. If it's type 3, we'll switch values
1323              * later.
1324              */
1325             $tagtype = 1;
1326             break;
1327     }
1328
1329     $tag_start = $pos;
1330     $tagname = '';
1331     /**
1332      * Look for next [\W-_], which will indicate the end of the tag name.
1333      */
1334     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1335     if ($regary == false){
1336         return Array(false, false, false, $lt, strlen($body));
1337     }
1338     list($pos, $tagname, $match) = $regary;
1339     $tagname = strtolower($tagname);
1340
1341     /**
1342      * $match can be either of these:
1343      * '>'  indicating the end of the tag entirely.
1344      * '\s' indicating the end of the tag name.
1345      * '/'  indicating that this is type-3 xhtml tag.
1346      *
1347      * Whatever else we find there indicates an invalid tag.
1348      */
1349     switch ($match){
1350         case '/':
1351             /**
1352              * This is an xhtml-style tag with a closing / at the
1353              * end, like so: <img src="blah" />. Check if it's followed
1354              * by the closing bracket. If not, then this tag is invalid
1355              */
1356             if (substr($body, $pos, 2) == "/>"){
1357                 $pos++;
1358                 $tagtype = 3;
1359             } else {
1360                 $gt = sq_findnxstr($body, $pos, ">");
1361                 $retary = Array(false, false, false, $lt, $gt);
1362                 return $retary;
1363             }
1364         case '>':
1365             return Array($tagname, false, $tagtype, $lt, $pos);
1366             break;
1367         default:
1368             /**
1369              * Check if it's whitespace
1370              */
1371             if (!preg_match('/\s/', $match)){
1372                 /**
1373                  * This is an invalid tag! Look for the next closing ">".
1374                  */
1375                 $gt = sq_findnxstr($body, $lt, ">");
1376                 return Array(false, false, false, $lt, $gt);
1377             }
1378             break;
1379     }
1380
1381     /**
1382      * At this point we're here:
1383      * <tagname  attribute='blah'>
1384      * \-------^
1385      *
1386      * At this point we loop in order to find all attributes.
1387      */
1388     $attname = '';
1389     $atttype = false;
1390     $attary = Array();
1391
1392     while ($pos <= strlen($body)){
1393         $pos = sq_skipspace($body, $pos);
1394         if ($pos == strlen($body)){
1395             /**
1396              * Non-closed tag.
1397              */
1398             return Array(false, false, false, $lt, $pos);
1399         }
1400         /**
1401          * See if we arrived at a ">" or "/>", which means that we reached
1402          * the end of the tag.
1403          */
1404         $matches = Array();
1405         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1406             /**
1407              * Yep. So we did.
1408              */
1409             $pos += strlen($matches{1});
1410             if ($matches{2} == "/>"){
1411                 $tagtype = 3;
1412                 $pos++;
1413             }
1414             return Array($tagname, $attary, $tagtype, $lt, $pos);
1415         }
1416
1417         /**
1418          * There are several types of attributes, with optional
1419          * [:space:] between members.
1420          * Type 1:
1421          *   attrname[:space:]=[:space:]'CDATA'
1422          * Type 2:
1423          *   attrname[:space:]=[:space:]"CDATA"
1424          * Type 3:
1425          *   attr[:space:]=[:space:]CDATA
1426          * Type 4:
1427          *   attrname
1428          *
1429          * We leave types 1 and 2 the same, type 3 we check for
1430          * '"' and convert to "&quot" if needed, then wrap in
1431          * double quotes. Type 4 we convert into:
1432          * attrname="yes".
1433          */
1434         $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1435         if ($regary == false){
1436             /**
1437              * Looks like body ended before the end of tag.
1438              */
1439             return Array(false, false, false, $lt, strlen($body));
1440         }
1441         list($pos, $attname, $match) = $regary;
1442         $attname = strtolower($attname);
1443         /**
1444          * We arrived at the end of attribute name. Several things possible
1445          * here:
1446          * '>'  means the end of the tag and this is attribute type 4
1447          * '/'  if followed by '>' means the same thing as above
1448          * '\s' means a lot of things -- look what it's followed by.
1449          *      anything else means the attribute is invalid.
1450          */
1451         switch($match){
1452             case '/':
1453                 /**
1454                  * This is an xhtml-style tag with a closing / at the
1455                  * end, like so: <img src="blah" />. Check if it's followed
1456                  * by the closing bracket. If not, then this tag is invalid
1457                  */
1458                 if (substr($body, $pos, 2) == "/>"){
1459                     $pos++;
1460                     $tagtype = 3;
1461                 } else {
1462                     $gt = sq_findnxstr($body, $pos, ">");
1463                     $retary = Array(false, false, false, $lt, $gt);
1464                     return $retary;
1465                 }
1466             case '>':
1467                 $attary{$attname} = '"yes"';
1468                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1469                 break;
1470             default:
1471                 /**
1472                  * Skip whitespace and see what we arrive at.
1473                  */
1474                 $pos = sq_skipspace($body, $pos);
1475                 $char = substr($body, $pos, 1);
1476                 /**
1477                  * Two things are valid here:
1478                  * '=' means this is attribute type 1 2 or 3.
1479                  * \w means this was attribute type 4.
1480                  * anything else we ignore and re-loop. End of tag and
1481                  * invalid stuff will be caught by our checks at the beginning
1482                  * of the loop.
1483                  */
1484                 if ($char == "="){
1485                     $pos++;
1486                     $pos = sq_skipspace($body, $pos);
1487                     /**
1488                      * Here are 3 possibilities:
1489                      * "'"  attribute type 1
1490                      * '"'  attribute type 2
1491                      * everything else is the content of tag type 3
1492                      */
1493                     $quot = substr($body, $pos, 1);
1494                     if ($quot == "'"){
1495                         $regary = sq_findnxreg($body, $pos+1, "\'");
1496                         if ($regary == false){
1497                             return Array(false, false, false, $lt, strlen($body));
1498                         }
1499                         list($pos, $attval, $match) = $regary;
1500                         $pos++;
1501                         $attary{$attname} = "'" . $attval . "'";
1502                     } else if ($quot == '"'){
1503                         $regary = sq_findnxreg($body, $pos+1, '\"');
1504                         if ($regary == false){
1505                             return Array(false, false, false, $lt, strlen($body));
1506                         }
1507                         list($pos, $attval, $match) = $regary;
1508                         $pos++;
1509                         $attary{$attname} = '"' . $attval . '"';
1510                     } else {
1511                         /**
1512                          * These are hateful. Look for \s, or >.
1513                          */
1514                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1515                         if ($regary == false){
1516                             return Array(false, false, false, $lt, strlen($body));
1517                         }
1518                         list($pos, $attval, $match) = $regary;
1519                         /**
1520                          * If it's ">" it will be caught at the top.
1521                          */
1522                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1523                         $attary{$attname} = '"' . $attval . '"';
1524                     }
1525                 } else if (preg_match("|[\w/>]|", $char)) {
1526                     /**
1527                      * That was attribute type 4.
1528                      */
1529                     $attary{$attname} = '"yes"';
1530                 } else {
1531                     /**
1532                      * An illegal character. Find next '>' and return.
1533                      */
1534                     $gt = sq_findnxstr($body, $pos, ">");
1535                     return Array(false, false, false, $lt, $gt);
1536                 }
1537                 break;
1538         }
1539     }
1540     /**
1541      * The fact that we got here indicates that the tag end was never
1542      * found. Return invalid tag indication so it gets stripped.
1543      */
1544     return Array(false, false, false, $lt, strlen($body));
1545 }
1546
1547 /**
1548  * Translates entities into literal values so they can be checked.
1549  *
1550  * @param $attvalue the by-ref value to check.
1551  * @param $regex    the regular expression to check against.
1552  * @param $hex      whether the entites are hexadecimal.
1553  * @return          True or False depending on whether there were matches.
1554  */
1555 function sq_deent(&$attvalue, $regex, $hex=false){
1556     $me = 'sq_deent';
1557     $ret_match = false;
1558     preg_match_all($regex, $attvalue, $matches);
1559     if (is_array($matches) && sizeof($matches[0]) > 0){
1560         $repl = Array();
1561         for ($i = 0; $i < sizeof($matches[0]); $i++){
1562             $numval = $matches[1][$i];
1563             if ($hex){
1564                 $numval = hexdec($numval);
1565             }
1566             $repl{$matches[0][$i]} = chr($numval);
1567         }
1568         $attvalue = strtr($attvalue, $repl);
1569         return true;
1570     } else {
1571         return false;
1572     }
1573 }
1574
1575 /**
1576  * This function runs various checks against the attributes.
1577  *
1578  * @param  $tagname         String with the name of the tag.
1579  * @param  $attary          Array with all tag attributes.
1580  * @param  $rm_attnames     See description for sq_sanitize
1581  * @param  $bad_attvals     See description for sq_sanitize
1582  * @param  $add_attr_to_tag See description for sq_sanitize
1583  * @param  $message         message object
1584  * @param  $id              message id
1585  * @return                  Array with modified attributes.
1586  */
1587 function sq_fixatts($tagname,
1588                     $attary,
1589                     $rm_attnames,
1590                     $bad_attvals,
1591                     $add_attr_to_tag,
1592                     $message,
1593                     $id,
1594                     $mailbox
1595                     ){
1596     $me = 'sq_fixatts';
1597     while (list($attname, $attvalue) = each($attary)){
1598         /**
1599          * See if this attribute should be removed.
1600          */
1601         foreach ($rm_attnames as $matchtag=>$matchattrs){
1602             if (preg_match($matchtag, $tagname)){
1603                 foreach ($matchattrs as $matchattr){
1604                     if (preg_match($matchattr, $attname)){
1605                         unset($attary{$attname});
1606                         continue;
1607                     }
1608                 }
1609             }
1610         }
1611         /**
1612          * Remove any backslashes, entities, and extraneous whitespace.
1613          */
1614         sq_defang($attvalue);
1615         sq_unspace($attvalue);
1616
1617         /**
1618          * Now let's run checks on the attvalues.
1619          * I don't expect anyone to comprehend this. If you do,
1620          * get in touch with me so I can drive to where you live and
1621          * shake your hand personally. :)
1622          */
1623         foreach ($bad_attvals as $matchtag=>$matchattrs){
1624             if (preg_match($matchtag, $tagname)){
1625                 foreach ($matchattrs as $matchattr=>$valary){
1626                     if (preg_match($matchattr, $attname)){
1627                         /**
1628                          * There are two arrays in valary.
1629                          * First is matches.
1630                          * Second one is replacements
1631                          */
1632                         list($valmatch, $valrepl) = $valary;
1633                         $newvalue =
1634                             preg_replace($valmatch, $valrepl, $attvalue);
1635                         if ($newvalue != $attvalue){
1636                             $attary{$attname} = $newvalue;
1637                         }
1638                     }
1639                 }
1640             }
1641         }
1642
1643         /**
1644          * Replace empty src tags with the blank image.  src is only used
1645          * for frames, images, and image inputs.  Doing a replace should
1646          * not affect them working as should be, however it will stop
1647          * IE from being kicked off when src for img tags are not set
1648          */
1649         if (($attname == 'src') && ($attvalue == '""')) {
1650             $attary{$attname} = '"' . SM_PATH . 'images/blank.png"';
1651         }
1652
1653         /**
1654          * Turn cid: urls into http-friendly ones.
1655          */
1656         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1657             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1658         }
1659
1660         /**
1661          * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1662          * One day MS might actually make it match something useful, for now, falling
1663          * back to using cid2http, so we can grab the blank.png.
1664          */
1665         if (preg_match("/^[\'\"]\s*outbind:\/\//si", $attvalue)) {
1666             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1667         }
1668
1669     }
1670     /**
1671      * See if we need to append any attributes to this tag.
1672      */
1673     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1674         if (preg_match($matchtag, $tagname)){
1675             $attary = array_merge($attary, $addattary);
1676         }
1677     }
1678     return $attary;
1679 }
1680
1681 /**
1682  * This function edits the style definition to make them friendly and
1683  * usable in SquirrelMail.
1684  *
1685  * @param  $message  the message object
1686  * @param  $id       the message id
1687  * @param  $content  a string with whatever is between <style> and </style>
1688  * @param  $mailbox  the message mailbox
1689  * @return           a string with edited content.
1690  */
1691 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1692     global $view_unsafe_images;
1693     $me = 'sq_fixstyle';
1694     $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1695     if ($ret == FALSE){
1696         return array(FALSE, strlen($body));
1697     }
1698     $newpos = $ret[0] + strlen($ret[2]);
1699     $content = $ret[1];
1700     /**
1701     * First look for general BODY style declaration, which would be
1702     * like so:
1703     * body {background: blah-blah}
1704     * and change it to .bodyclass so we can just assign it to a <div>
1705     */
1706     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1707     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1708     /**
1709     * Fix url('blah') declarations.
1710     */
1711     //   $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1712     //                           "url(\\1$secremoveimg\\2)", $content);
1713     // remove NUL
1714     $content = str_replace("\0", "", $content);
1715     // translate ur\l and variations (IE parses that)
1716     $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
1717     // NB I insert NUL characters to keep to avoid an infinite loop. They are removed after the loop.
1718     while (preg_match("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si", $content, $matches)) {
1719         $sProto = strtolower($matches[1]);
1720         switch ($sProto) {
1721             /**
1722              * Fix url('https*://.*) declarations but only if $view_unsafe_images
1723              * is false.
1724              */
1725             case 'https':
1726             case 'http':
1727                 if (!$view_unsafe_images){
1728
1729                     $sExpr = "/url\s*\(\s*[\'\"]?\s*$sProto*:.*[\'\"]?\s*\)/si";
1730                     $content = preg_replace($sExpr, "u\0r\0l(\\1$secremoveimg\\2)", $content);
1731
1732                 } else {
1733                     $content = preg_replace('/url/i',"u\0r\0l",$content);
1734                 }
1735                 break;
1736             /**
1737              * Fix urls that refer to cid:
1738              */
1739             case 'cid':
1740                 $cidurl = 'cid:'. $matches[2];
1741                 $httpurl = sq_cid2http($message, $id, $cidurl, $mailbox);
1742                 // escape parentheses that can modify the regular expression
1743                 $cidurl = str_replace(array('(',')'),array('\\(','\\)'),$cidurl);
1744                 $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1745                                         "u\0r\0l($httpurl)", $content);
1746                 break;
1747             default:
1748                 /**
1749                  * replace url with protocol other then the white list
1750                  * http,https and cid by an empty string.
1751                  */
1752                 $content = preg_replace("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si",
1753                                  "", $content);
1754                 break;
1755         }
1756     }
1757     // remove NUL
1758     $content = str_replace("\0", "", $content);
1759    /**
1760     * Remove any backslashes, entities, and extraneous whitespace.
1761     */
1762     $contentTemp = $content;
1763     sq_defang($contentTemp);
1764     sq_unspace($contentTemp);
1765
1766     /**
1767      * Fix stupid css declarations which lead to vulnerabilities
1768      * in IE.
1769      */
1770     $match   = Array('/\/\*.*\*\//',
1771                     '/expression/i',
1772                     '/behaviou*r/i',
1773                     '/binding/i',
1774                     '/include-source/i');
1775     $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy');
1776     $contentNew = preg_replace($match, $replace, $contentTemp);
1777     if ($contentNew !== $contentTemp) {
1778         // insecure css declarations are used. From now on we don't care
1779         // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
1780         $content = $contentNew;
1781     }
1782     return array($content, $newpos);
1783 }
1784
1785
1786 /**
1787  * This function converts cid: url's into the ones that can be viewed in
1788  * the browser.
1789  *
1790  * @param  $message  the message object
1791  * @param  $id       the message id
1792  * @param  $cidurl   the cid: url.
1793  * @param  $mailbox  the message mailbox
1794  * @return           a string with a http-friendly url
1795  */
1796 function sq_cid2http($message, $id, $cidurl, $mailbox){
1797     /**
1798      * Get rid of quotes.
1799      */
1800     $quotchar = substr($cidurl, 0, 1);
1801     if ($quotchar == '"' || $quotchar == "'"){
1802         $cidurl = str_replace($quotchar, "", $cidurl);
1803     } else {
1804         $quotchar = '';
1805     }
1806     $cidurl = substr(trim($cidurl), 4);
1807
1808     $match_str = '/\{.*?\}\//';
1809     $str_rep = '';
1810     $cidurl = preg_replace($match_str, $str_rep, $cidurl);
1811
1812     $linkurl = find_ent_id($cidurl, $message);
1813     /* in case of non-save cid links $httpurl should be replaced by a sort of
1814        unsave link image */
1815     $httpurl = '';
1816
1817     /**
1818      * This is part of a fix for Outlook Express 6.x generating
1819      * cid URLs without creating content-id headers. These images are
1820      * not part of the multipart/related html mail. The html contains
1821      * <img src="cid:{some_id}/image_filename.ext"> references to
1822      * attached images with as goal to render them inline although
1823      * the attachment disposition property is not inline.
1824      */
1825
1826     if (empty($linkurl)) {
1827         if (preg_match('/{.*}\//', $cidurl)) {
1828             $cidurl = preg_replace('/{.*}\//','', $cidurl);
1829             if (!empty($cidurl)) {
1830                 $linkurl = find_ent_id($cidurl, $message);
1831             }
1832         }
1833     }
1834
1835     if (!empty($linkurl)) {
1836         $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1837             "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1838             '&amp;ent_id=' . $linkurl . $quotchar;
1839     } else {
1840         /**
1841          * If we couldn't generate a proper img url, drop in a blank image
1842          * instead of sending back empty, otherwise it causes unusual behaviour
1843          */
1844         $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
1845     }
1846
1847     return $httpurl;
1848 }
1849
1850 /**
1851  * This function changes the <body> tag into a <div> tag since we
1852  * can't really have a body-within-body.
1853  *
1854  * @param  $attary   an array of attributes and values of <body>
1855  * @param  $mailbox  mailbox we're currently reading (for cid2http)
1856  * @param  $message  current message (for cid2http)
1857  * @param  $id       current message id (for cid2http)
1858  * @return           a modified array of attributes to be set for <div>
1859  */
1860 function sq_body2div($attary, $mailbox, $message, $id){
1861     $me = 'sq_body2div';
1862     $divattary = Array('class' => "'bodyclass'");
1863     $text = '#000000';
1864     $has_bgc_stl = $has_txt_stl = false;
1865     $styledef = '';
1866     if (is_array($attary) && sizeof($attary) > 0){
1867         foreach ($attary as $attname=>$attvalue){
1868             $quotchar = substr($attvalue, 0, 1);
1869             $attvalue = str_replace($quotchar, "", $attvalue);
1870             switch ($attname){
1871                 case 'background':
1872                     $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1873                     $styledef .= "background-image: url('$attvalue'); ";
1874                     break;
1875                 case 'bgcolor':
1876                     $has_bgc_stl = true;
1877                     $styledef .= "background-color: $attvalue; ";
1878                     break;
1879                 case 'text':
1880                     $has_txt_stl = true;
1881                     $styledef .= "color: $attvalue; ";
1882                     break;
1883             }
1884         }
1885         // Outlook defines a white bgcolor and no text color. This can lead to
1886         // white text on a white bg with certain themes.
1887         if ($has_bgc_stl && !$has_txt_stl) {
1888             $styledef .= "color: $text; ";
1889         }
1890         if (strlen($styledef) > 0){
1891             $divattary{"style"} = "\"$styledef\"";
1892         }
1893     }
1894     return $divattary;
1895 }
1896
1897 /**
1898  * This is the main function and the one you should actually be calling.
1899  * There are several variables you should be aware of an which need
1900  * special description.
1901  *
1902  * Since the description is quite lengthy, see it here:
1903  * http://linux.duke.edu/projects/mini/htmlfilter/
1904  *
1905  * @param $body                 the string with HTML you wish to filter
1906  * @param $tag_list             see description above
1907  * @param $rm_tags_with_content see description above
1908  * @param $self_closing_tags    see description above
1909  * @param $force_tag_closing    see description above
1910  * @param $rm_attnames          see description above
1911  * @param $bad_attvals          see description above
1912  * @param $add_attr_to_tag      see description above
1913  * @param $message              message object
1914  * @param $id                   message id
1915  * @return                      sanitized html safe to show on your pages.
1916  */
1917 function sq_sanitize($body,
1918                      $tag_list,
1919                      $rm_tags_with_content,
1920                      $self_closing_tags,
1921                      $force_tag_closing,
1922                      $rm_attnames,
1923                      $bad_attvals,
1924                      $add_attr_to_tag,
1925                      $message,
1926                      $id,
1927                      $mailbox
1928                      ){
1929     $me = 'sq_sanitize';
1930     $rm_tags = array_shift($tag_list);
1931     /**
1932      * Normalize rm_tags and rm_tags_with_content.
1933      */
1934     @array_walk($tag_list, 'sq_casenormalize');
1935     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1936     @array_walk($self_closing_tags, 'sq_casenormalize');
1937     /**
1938      * See if tag_list is of tags to remove or tags to allow.
1939      * false  means remove these tags
1940      * true   means allow these tags
1941      */
1942     $curpos = 0;
1943     $open_tags = Array();
1944     $trusted = "\n<!-- begin sanitized html -->\n";
1945     $skip_content = false;
1946     /**
1947      * Take care of netscape's stupid javascript entities like
1948      * &{alert('boo')};
1949      */
1950     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1951
1952     while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1953         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1954         $free_content = substr($body, $curpos, $lt-$curpos);
1955         /**
1956          * Take care of <style>
1957          */
1958         if ($tagname == "style" && $tagtype == 1){
1959             list($free_content, $curpos) =
1960                 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
1961             if ($free_content != FALSE){
1962                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1963                 $trusted .= $free_content;
1964                 $trusted .= sq_tagprint($tagname, false, 2);
1965             }
1966             continue;
1967         }
1968         if ($skip_content == false){
1969             $trusted .= $free_content;
1970         }
1971         if ($tagname != FALSE){
1972             if ($tagtype == 2){
1973                 if ($skip_content == $tagname){
1974                     /**
1975                      * Got to the end of tag we needed to remove.
1976                      */
1977                     $tagname = false;
1978                     $skip_content = false;
1979                 } else {
1980                     if ($skip_content == false){
1981                         if ($tagname == "body"){
1982                             $tagname = "div";
1983                         }
1984                         if (isset($open_tags{$tagname}) &&
1985                                 $open_tags{$tagname} > 0){
1986                             $open_tags{$tagname}--;
1987                         } else {
1988                             $tagname = false;
1989                         }
1990                     }
1991                 }
1992             } else {
1993                 /**
1994                  * $rm_tags_with_content
1995                  */
1996                 if ($skip_content == false){
1997                     /**
1998                      * See if this is a self-closing type and change
1999                      * tagtype appropriately.
2000                      */
2001                     if ($tagtype == 1
2002                             && in_array($tagname, $self_closing_tags)){
2003                         $tagtype = 3;
2004                     }
2005                     /**
2006                      * See if we should skip this tag and any content
2007                      * inside it.
2008                      */
2009                     if ($tagtype == 1 &&
2010                             in_array($tagname, $rm_tags_with_content)){
2011                         $skip_content = $tagname;
2012                     } else {
2013                         if (($rm_tags == false
2014                                     && in_array($tagname, $tag_list)) ||
2015                                 ($rm_tags == true &&
2016                                  !in_array($tagname, $tag_list))){
2017                             $tagname = false;
2018                         } else {
2019                             /**
2020                              * Convert body into div.
2021                              */
2022                             if ($tagname == "body"){
2023                                 $tagname = "div";
2024                                 $attary = sq_body2div($attary, $mailbox,
2025                                         $message, $id);
2026                             }
2027                             if ($tagtype == 1){
2028                                 if (isset($open_tags{$tagname})){
2029                                     $open_tags{$tagname}++;
2030                                 } else {
2031                                     $open_tags{$tagname}=1;
2032                                 }
2033                             }
2034                             /**
2035                              * This is where we run other checks.
2036                              */
2037                             if (is_array($attary) && sizeof($attary) > 0){
2038                                 $attary = sq_fixatts($tagname,
2039                                                      $attary,
2040                                                      $rm_attnames,
2041                                                      $bad_attvals,
2042                                                      $add_attr_to_tag,
2043                                                      $message,
2044                                                      $id,
2045                                                      $mailbox
2046                                                      );
2047                             }
2048                         }
2049                     }
2050                 }
2051             }
2052             if ($tagname != false && $skip_content == false){
2053                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2054             }
2055         }
2056         $curpos = $gt+1;
2057     }
2058     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2059     if ($force_tag_closing == true){
2060         foreach ($open_tags as $tagname=>$opentimes){
2061             while ($opentimes > 0){
2062                 $trusted .= '</' . $tagname . '>';
2063                 $opentimes--;
2064             }
2065         }
2066         $trusted .= "\n";
2067     }
2068     $trusted .= "<!-- end sanitized html -->\n";
2069     return $trusted;
2070 }
2071
2072 /**
2073  * This is a wrapper function to call html sanitizing routines.
2074  *
2075  * @param  $body  the body of the message
2076  * @param  $id    the id of the message
2077
2078  * @param  $message
2079  * @param  $mailbox
2080  * @param  boolean $take_mailto_links When TRUE, converts mailto: links
2081  *                                    into internal SM compose links
2082  *                                    (optional; default = TRUE)
2083  * @return        a string with html safe to display in the browser.
2084  */
2085 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2086
2087     // require_once(SM_PATH . 'functions/url_parser.php');  // for $MailTo_PReg_Match
2088
2089     global $attachment_common_show_images, $view_unsafe_images,
2090            $has_unsafe_images;
2091     /**
2092      * Don't display attached images in HTML mode.
2093      *
2094      * SB: why?
2095      */
2096     $attachment_common_show_images = false;
2097     $tag_list = Array(
2098             false,
2099             "object",
2100             "meta",
2101             "html",
2102             "head",
2103             "base",
2104             "link",
2105             "frame",
2106             "iframe",
2107             "plaintext",
2108             "marquee"
2109             );
2110
2111     $rm_tags_with_content = Array(
2112             "script",
2113             "applet",
2114             "embed",
2115             "title",
2116             "frameset",
2117             "xmp",
2118             "xml"
2119             );
2120
2121     $self_closing_tags =  Array(
2122             "img",
2123             "br",
2124             "hr",
2125             "input",
2126             "outbind"
2127             );
2128
2129     $force_tag_closing = true;
2130
2131     $rm_attnames = Array(
2132             "/.*/" =>
2133             Array(
2134                 "/target/i",
2135                 "/^on.*/i",
2136                 "/^dynsrc/i",
2137                 "/^data.*/i",
2138                 "/^lowsrc.*/i"
2139                 )
2140             );
2141
2142     $secremoveimg = "../images/" . _("sec_remove_eng.png");
2143     $bad_attvals = Array(
2144             "/.*/" =>
2145             Array(
2146                 "/^src|background/i" =>
2147                 Array(
2148                     Array(
2149                         "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2150                         "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2151                         "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2152                         ),
2153                     Array(
2154                         "\\1$secremoveimg\\2",
2155                         "\\1$secremoveimg\\2",
2156                         "\\1$secremoveimg\\2",
2157                         )
2158                     ),
2159                 "/^href|action/i" =>
2160                 Array(
2161                     Array(
2162                         "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2163                         "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2164                         "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2165                         ),
2166                     Array(
2167                         "\\1#\\1",
2168                         "\\1#\\1",
2169                         "\\1#\\1"
2170                         )
2171                     ),
2172         "/^style/i" =>
2173             Array(
2174                 Array(
2175                     "/\/\*.*\*\//",
2176                     "/expression/i",
2177                     "/binding/i",
2178                     "/behaviou*r/i",
2179                     "/include-source/i",
2180                     "/position\s*:\s*absolute/i",
2181                     "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
2182                     "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2183                     "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2184                     "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2185                     "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2186                     ),
2187                 Array(
2188                     "",
2189                     "idiocy",
2190                     "idiocy",
2191                     "idiocy",
2192                     "idiocy",
2193                     "",
2194                     "url",
2195                     "url(\\1#\\1)",
2196                     "url(\\1#\\1)",
2197                     "url(\\1#\\1)",
2198                     "\\1:url(\\2#\\3)"
2199                     )
2200                 )
2201             )
2202         );
2203     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
2204         $view_unsafe_images = false;
2205     }
2206     if (!$view_unsafe_images){
2207         /**
2208          * Remove any references to http/https if view_unsafe_images set
2209          * to false.
2210          */
2211         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2212                 '/^([\'\"])\s*https*:.*([\'\"])/si');
2213         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2214                 "\\1$secremoveimg\\1");
2215         array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2216                 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2217         array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2218                 "url(\\1$secremoveimg\\1)");
2219     }
2220
2221     $add_attr_to_tag = Array(
2222             "/^a$/i" =>
2223             Array('target'=>'"_blank"',
2224                 'title'=>'"'._("This external link will open in a new window").'"'
2225                 )
2226             );
2227     $trusted = sq_sanitize($body,
2228                            $tag_list,
2229                            $rm_tags_with_content,
2230                            $self_closing_tags,
2231                            $force_tag_closing,
2232                            $rm_attnames,
2233                            $bad_attvals,
2234                            $add_attr_to_tag,
2235                            $message,
2236                            $id,
2237                            $mailbox
2238                            );
2239     if (preg_match("|$secremoveimg|i", $trusted)){
2240         $has_unsafe_images = true;
2241     }
2242
2243     // we want to parse mailto's in HTML output, change to SM compose links
2244     // this is a modified version of code from url_parser.php... but Marc is
2245     // right: we need a better filtering implementation; adding this randomly
2246     // here is not a great solution
2247     //
2248     if ($take_mailto_links) {
2249         // parseUrl($trusted);   // this even parses URLs inside of tags... too aggressive
2250         global $MailTo_PReg_Match;
2251         $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1) ;
2252         if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2253             foreach ($regs[0] as $i => $mailto_before) {
2254                 $mailto_params = $regs[10][$i];
2255                 // get rid of any tailing quote since we have to add send_to to the end
2256                 //
2257                 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2258                     $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2259                 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2260                     $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2261
2262                 if ($regs[1][$i]) {    //if there is an email addr before '?', we need to merge it with the params
2263                     $to = 'to=' . $regs[1][$i];
2264                     if (strpos($mailto_params, 'to=') > -1)    //already a 'to='
2265                         $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2266                     else {
2267                         if ($mailto_params)    //already some params, append to them
2268                             $mailto_params .= '&amp;' . $to;
2269                         else
2270                             $mailto_params .= '?' . $to;
2271                     }
2272                 }
2273
2274                 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2275
2276                 // we'll already have target=_blank, no need to allow comp_in_new
2277                 // here (which would be a lot more work anyway)
2278                 //
2279                 global $compose_new_win;
2280                 $temp_comp_in_new = $compose_new_win;
2281                 $compose_new_win = 0;
2282                 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2283                 $compose_new_win = $temp_comp_in_new;
2284
2285                 // remove <a href=" and anything after the next quote (we only
2286                 // need the uri, not the link HTML) in compose uri
2287                 //
2288                 $comp_uri = substr($comp_uri, 9);
2289                 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2290                 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2291             }
2292         }
2293     }
2294
2295     return $trusted;
2296 }
2297
2298 /**
2299  * function SendDownloadHeaders - send file to the browser
2300  *
2301  * Original Source: SM core src/download.php
2302  * moved here to make it available to other code, and separate
2303  * front end from back end functionality.
2304  *
2305  * @param string $type0 first half of mime type
2306  * @param string $type1 second half of mime type
2307  * @param string $filename filename to tell the browser for downloaded file
2308  * @param boolean $force whether to force the download dialog to pop
2309  * @param optional integer $filesize send the Content-Header and length to the browser
2310  * @return void
2311  */
2312 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2313     global $languages, $squirrelmail_language;
2314     $isIE = $isIE6plus = false;
2315
2316     sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2317
2318     if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2319             strstr($HTTP_USER_AGENT, 'Opera') === false) {
2320         $isIE = true;
2321     }
2322
2323     if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2324         ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2325         $isIE6plus = true;
2326     }
2327
2328     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2329             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2330         $filename =
2331             call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2332     } else {
2333         $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
2334     }
2335
2336     // A Pox on Microsoft and it's Internet Explorer!
2337     //
2338     // IE has lots of bugs with file downloads.
2339     // It also has problems with SSL.  Both of these cause problems
2340     // for us in this function.
2341     //
2342     // See this article on Cache Control headers and SSL
2343     // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2344     //
2345     // The best thing you can do for IE is to upgrade to the latest
2346     // version
2347     //set all the Cache Control Headers for IE
2348     if ($isIE) {
2349         $filename=rawurlencode($filename);
2350         header ("Pragma: public");
2351         header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2352         header ("Cache-Control: post-check=0, pre-check=0", false);
2353         header ("Cache-Control: private");
2354
2355         //set the inline header for IE, we'll add the attachment header later if we need it
2356         header ("Content-Disposition: inline; filename=$filename");
2357     }
2358
2359     if (!$force) {
2360         // Try to show in browser window
2361         header ("Content-Disposition: inline; filename=\"$filename\"");
2362         header ("Content-Type: $type0/$type1; name=\"$filename\"");
2363     } else {
2364         // Try to pop up the "save as" box
2365
2366         // IE makes this hard.  It pops up 2 save boxes, or none.
2367         // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2368         // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2369         // But, according to Microsoft, it is "RFC compliant but doesn't
2370         // take into account some deviations that allowed within the
2371         // specification."  Doesn't that mean RFC non-compliant?
2372         // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2373
2374         // all browsers need the application/octet-stream header for this
2375         header ("Content-Type: application/octet-stream; name=\"$filename\"");
2376
2377         // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2378         // Do not have quotes around filename, but that applied to
2379         // "attachment"... does it apply to inline too?
2380         header ("Content-Disposition: attachment; filename=\"$filename\"");
2381
2382         if ($isIE && !$isIE6plus) {
2383             // This combination seems to work mostly.  IE 5.5 SP 1 has
2384             // known issues (see the Microsoft Knowledge Base)
2385
2386             // This works for most types, but doesn't work with Word files
2387             header ("Content-Type: application/download; name=\"$filename\"");
2388
2389             // These are spares, just in case.  :-)
2390             //header("Content-Type: $type0/$type1; name=\"$filename\"");
2391             //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2392             //header("Content-Type: application/octet-stream; name=\"$filename\"");
2393         } else {
2394             // another application/octet-stream forces download for Netscape
2395             header ("Content-Type: application/octet-stream; name=\"$filename\"");
2396         }
2397     }
2398
2399     //send the content-length header if the calling function provides it
2400     if ($filesize > 0) {
2401         header("Content-Length: $filesize");
2402     }
2403
2404 }  // end fn SendDownloadHeaders