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                 // adding warning message
 418                 $body = html_tag('div',_("Viewing HTML formatted email"),'center');
 419
 420                 /**
 421                  * height can't be set to 100%, because it does not work as expected when
 422                  * iframe is inside the table. Browsers do not create full height objects
 423                  * even when iframe is not nested. Maybe there is some way to get full size
 424                  * with CSS. Tested in firefox 1.02 and opera 7.53
 425                  *
 426                  * width="100%" does not work as expected, when table width is not set (automatic)
 427                  *
 428                  * tokul: I think <iframe> are safer sandbox than <object>. Objects might
 429                  * need special handling for IE and IE6SP2.
 430                  */
 431                 $body.= "<div><iframe name=\"message_frame\" width=\"100%\" height=\"$iframe_height\" src=\"$iframeurl\""
 432                     .' frameborder="1" marginwidth="0" marginheight="0" scrolling="auto">' . "\n";
 433
 434                 // Message for browsers without iframe support
 435                 //$body.= _("Your browser does not support inline frames.
 436                 // You can view HTML formated message by following below link.");
 437                 //$body.= "<br /><a href=\"$iframeurl\">"._("View HTML Message")."</a>";
 438
 439                 // if browser can't render iframe, it renders html message.
 440                 $body.= $html_body;
 441
 442                 // close iframe
 443                 $body.="</iframe></div>\n";
 444             } else {
 445                 // old way of html rendering
 446                 $body = magicHTML($body, $id, $message, $mailbox);
 447                 /**
 448                  * convert character set. charset_decode does not remove html special chars
 449                  * applied by magicHTML functions and does not sanitize them second time if
 450                  * fourth argument is true.
 451                  */
 452                 $body = charset_decode($body_message->header->getParameter('charset'),$body,false,true);
 453             }
 454         } else {
 455             translateText($body, $wrap_at,
 456                     $body_message->header->getParameter('charset'));
 457         }
 458
 459         // if this is the clean display (i.e. printer friendly), stop here.
 460         if ( $clean ) {
 461             return $body;
 462         }
 463
 464         $download_and_unsafe_link = '';
 465
 466         $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
 467             '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
 468             '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 469         if (isset($passed_ent_id)) {
 470             $link .= '&amp;passed_ent_id='.$passed_ent_id;
 471         }
 472         $download_href = SM_PATH . 'src/download.php?absolute_dl=true&amp;' . $link;
 473         $download_and_unsafe_link .= '&nbsp;|&nbsp;<a href="'. $download_href .'">' . _("Download this as a file") .  '</a>';
 474         if ($view_unsafe_images) {
 475             $text = _("Hide Unsafe Images");
 476         } else {
 477             if (isset($has_unsafe_images) && $has_unsafe_images) {
 478                 $link .= '&amp;view_unsafe_images=1';
 479                 $text = _("View Unsafe Images");
 480             } else {
 481                 $text = '';
 482             }
 483         }
 484         if($text != '') {
 485             $unsafe_image_toggle_href = SM_PATH . 'src/read_body.php?'.$link;
 486             $unsafe_image_toggle_text = $text;
 487             $download_and_unsafe_link .= '&nbsp;|&nbsp;<a href="'. $unsafe_image_toggle_href .'">' . $text . '</a>';
 488         }
 489     }
 490     return $body;
 491 }
 492
 493 /**
 494  * Generate attachments array for passing to templates.  Separated from
 495  * formatAttachments() below so that the same array can be given to the
 496  * print-friendly version.
 497  *
 498  * @since 1.5.2
 499  * @param object $message SquirrelMail message object
 500  * @param array $exclude_id message parts that are not attachments.
 501  * @param string $mailbox mailbox name
 502  * @param integer $id message id
 503  */
 504 function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
 505     global $where, $what, $startMessage, $color, $passed_ent_id, $base_uri;
 506
 507     $att_ar = $message->getAttachments($exclude_id);
 508     $urlMailbox = urlencode($mailbox);
 509
 510     $attachments = array();
 511     foreach ($att_ar as $att) {
 512         $ent = $att->entity_id;
 513         $header = $att->header;
 514         $type0 = strtolower($header->type0);
 515         $type1 = strtolower($header->type1);
 516         $name = '';
 517         $links = array();
 518         $links['download link']['text'] = _("Download");
 519         $links['download link']['href'] = $base_uri .
 520             "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 521
 522         if ($type0 =='message' && $type1 == 'rfc822') {
 523             $default_page = $base_uri  . 'src/read_body.php';
 524             $rfc822_header = $att->rfc822_header;
 525             $filename = $rfc822_header->subject;
 526             if (trim( $filename ) == '') {
 527                 $filename = 'untitled-[' . $ent . ']' ;
 528             }
 529             $from_o = $rfc822_header->from;
 530             if (is_object($from_o)) {
 531                 $from_name = decodeHeader($from_o->getAddress(false));
 532             } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
 533                 // something weird happens when a digest message is opened and you return to the digest
 534                 // now the from object is part of an array. Probably the parseHeader call overwrites the info
 535                 // retrieved from the bodystructure in a different way. We need to fix this later.
 536                 // possible starting point, do not fetch header we already have and inspect how
 537                 // the rfc822_header object behaves.
 538                 $from_name = decodeHeader($from_o[0]->getAddress(false));
 539             } else {
 540                 $from_name = _("Unknown sender");
 541             }
 542             $description = _("From").': '.$from_name;
 543         } else {
 544             $default_page = $base_uri  . 'src/download.php';
 545             $filename = $att->getFilename();
 546             if ($header->description) {
 547                 $description = decodeHeader($header->description);
 548             } else {
 549                 $description = '';
 550             }
 551         }
 552
 553         $display_filename = $filename;
 554         if (isset($passed_ent_id)) {
 555             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 556         } else {
 557             $passed_ent_id_link = '';
 558         }
 559         $defaultlink = $default_page . "?startMessage=$startMessage"
 560             . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 561             . '&amp;ent_id='.$ent.$passed_ent_id_link;
 562         if ($where && $what) {
 563             $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 564         }
 565
 566         /* This executes the attachment hook with a specific MIME-type.
 567          * If that doesn't have results, it tries if there's a rule
 568          * for a more generic type.
 569          */
 570         $hookresults = do_hook("attachment $type0/$type1", $links,
 571                 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 572                 $display_filename, $where, $what);
 573         if(count($hookresults[1]) <= 1) {
 574             $hookresults = do_hook("attachment $type0/*", $links,
 575                     $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 576                     $display_filename, $where, $what);
 577         }
 578
 579         $links = $hookresults[1];
 580         $defaultlink = $hookresults[6];
 581
 582         $this_attachment = array();
 583         $this_attachment['Name'] = decodeHeader($display_filename);
 584         $this_attachment['Description'] = $description;
 585         $this_attachment['DefaultHREF'] = $defaultlink;
 586         $this_attachment['DownloadHREF'] = $links['download link']['href'];
 587         $this_attachment['ViewHREF'] = isset($links['attachment_common']) ? $links['attachment_common']['href'] : '';
 588         $this_attachment['Size'] = $header->size;
 589         $this_attachment['ContentType'] = htmlspecialchars($type0 .'/'. $type1);
 590         $this_attachment['OtherLinks'] = array();
 591         foreach ($links as $val) {
 592             if ($val['text']==_("Download") || $val['text'] == _("View"))
 593                 continue;
 594             if (empty($val['text']) && empty($val['extra']))
 595                 continue;
 596
 597             $temp = array();
 598             $temp['HREF'] = $val['href'];
 599             $temp['Text'] = (empty($val['text']) ? '' : $val['text']) . (empty($val['extra']) ? '' : $val['extra']);
 600             $this_attachment['OtherLinks'][] = $temp;
 601         }
 602         $attachments[] = $this_attachment;
 603
 604         unset($links);
 605     }
 606
 607     return $attachments;
 608 }
 609
 610 /**
 611  * Displays attachment links and information
 612  *
 613  * Since 1.3.0 function is not included in formatBody() call.
 614  *
 615  * Since 1.0.2 uses attachment $type0/$type1 hook.
 616  * Since 1.2.5 uses attachment $type0/* hook.
 617  * Since 1.5.0 uses attachments_bottom hook.
 618  * Since 1.5.2 uses templates and does *not* return a value.
 619  *
 620  * @param object $message SquirrelMail message object
 621  * @param array $exclude_id message parts that are not attachments.
 622  * @param string $mailbox mailbox name
 623  * @param integer $id message id
 624  */
 625 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 626     global $oTemplate;
 627
 628     $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
 629
 630     $oTemplate->assign('attachments', $attach);
 631     $oTemplate->display('read_attachments.tpl');
 632 }
 633
 634 function sqimap_base64_decode(&$string) {
 635
 636     // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
 637     // fly decoding (to reduce memory usage) you have to check if the
 638     // data has incomplete pairs
 639
 640     // Remove the noise in order to check if the 4 bytes pairs are complete
 641     $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
 642
 643     $sStringRem = '';
 644     $iMod = strlen($string) % 4;
 645     if ($iMod) {
 646         $sStringRem = substr($string,-$iMod);
 647         // Check if $sStringRem contains padding characters
 648         if (substr($sStringRem,-1) != '=') {
 649             $string = substr($string,0,-$iMod);
 650         } else {
 651             $sStringRem = '';
 652         }
 653     }
 654     $string = base64_decode($string);
 655     return $sStringRem;
 656 }
 657
 658 /**
 659  * Decodes encoded message body
 660  *
 661  * This function decodes the body depending on the encoding type.
 662  * Currently quoted-printable and base64 encodings are supported.
 663  * decode_body hook was added to this function in 1.4.2/1.5.0
 664  * @param string $body encoded message body
 665  * @param string $encoding used encoding
 666  * @return string decoded string
 667  * @since 1.0
 668  */
 669 function decodeBody($body, $encoding) {
 670
 671     $body = str_replace("\r\n", "\n", $body);
 672     $encoding = strtolower($encoding);
 673
 674     $encoding_handler = do_hook_function('decode_body', $encoding);
 675
 676
 677     // plugins get first shot at decoding the body
 678     //
 679     if (!empty($encoding_handler) && function_exists($encoding_handler)) {
 680         $body = $encoding_handler('decode', $body);
 681
 682     } elseif ($encoding == 'quoted-printable' ||
 683             $encoding == 'quoted_printable') {
 684         /**
 685          * quoted_printable_decode() function is broken in older
 686          * php versions. Text with \r\n decoding was fixed only
 687          * in php 4.3.0. Minimal code requirement 4.0.4 +
 688          * str_replace("\r\n", "\n", $body); call.
 689          */
 690         $body = quoted_printable_decode($body);
 691     } elseif ($encoding == 'base64') {
 692         $body = base64_decode($body);
 693     }
 694
 695     // All other encodings are returned raw.
 696     return $body;
 697 }
 698
 699 /**
 700  * Decodes headers
 701  *
 702  * This functions decode strings that is encoded according to
 703  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 704  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 705  *
 706  * @param string $string header string that has to be made readable
 707  * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
 708  * @param boolean $htmlsave preserve spaces and sanitize html special characters. defaults to true
 709  * @param boolean $decide decide if string can be utfencoded. defaults to false
 710  * @return string decoded header string
 711  */
 712 function decodeHeader ($string, $utfencode=true,$htmlsave=true,$decide=false) {
 713     global $languages, $squirrelmail_language,$default_charset;
 714     if (is_array($string)) {
 715         $string = implode("\n", $string);
 716     }
 717
 718     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 719             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
 720         $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
 721         // Do we need to return at this point?
 722         // return $string;
 723     }
 724     $i = 0;
 725     $iLastMatch = -2;
 726     $encoded = true;
 727
 728     $aString = explode(' ',$string);
 729     $ret = '';
 730     foreach ($aString as $chunk) {
 731         if ($encoded && $chunk === '') {
 732             continue;
 733         } elseif ($chunk === '') {
 734             $ret .= ' ';
 735             continue;
 736         }
 737         $encoded = false;
 738         /* if encoded words are not separated by a linear-space-white we still catch them */
 739         $j = $i-1;
 740
 741         while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 742             /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
 743             if ($iLastMatch !== $j) {
 744                 if ($htmlsave) {
 745                     $ret .= '&#32;';
 746                 } else {
 747                     $ret .= ' ';
 748                 }
 749             }
 750             $iLastMatch = $i;
 751             $j = $i;
 752             if ($htmlsave) {
 753                 $ret .= htmlspecialchars($res[1]);
 754             } else {
 755                 $ret .= $res[1];
 756             }
 757             $encoding = ucfirst($res[3]);
 758
 759             /* decide about valid decoding */
 760             if ($decide && is_conversion_safe($res[2])) {
 761                 $utfencode=true;
 762                 $can_be_encoded=true;
 763             } else {
 764                 $can_be_encoded=false;
 765             }
 766             switch ($encoding)
 767             {
 768                 case 'B':
 769                     $replace = base64_decode($res[4]);
 770                     if ($utfencode) {
 771                         if ($can_be_encoded) {
 772                             /* convert string to different charset,
 773                              * if functions asks for it (usually in compose)
 774                              */
 775                             $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsave);
 776                         } else {
 777                             // convert string to html codes in order to display it
 778                             $ret .= charset_decode($res[2],$replace);
 779                         }
 780                     } else {
 781                         if ($htmlsave) {
 782                             $replace = htmlspecialchars($replace);
 783                         }
 784                         $ret.= $replace;
 785                     }
 786                     break;
 787                 case 'Q':
 788                     $replace = str_replace('_', ' ', $res[4]);
 789                     $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 790                             $replace);
 791                     if ($utfencode) {
 792                         if ($can_be_encoded) {
 793                             /* convert string to different charset,
 794                              * if functions asks for it (usually in compose)
 795                              */
 796                             $replace = charset_convert($res[2], $replace,$default_charset,$htmlsave);
 797                         } else {
 798                             // convert string to html codes in order to display it
 799                             $replace = charset_decode($res[2], $replace);
 800                         }
 801                     } else {
 802                         if ($htmlsave) {
 803                             $replace = htmlspecialchars($replace);
 804                         }
 805                     }
 806                     $ret .= $replace;
 807                     break;
 808                 default:
 809                     break;
 810             }
 811             $chunk = $res[5];
 812             $encoded = true;
 813         }
 814         if (!$encoded) {
 815             if ($htmlsave) {
 816                 $ret .= '&#32;';
 817             } else {
 818                 $ret .= ' ';
 819             }
 820         }
 821
 822         if (!$encoded && $htmlsave) {
 823             $ret .= htmlspecialchars($chunk);
 824         } else {
 825             $ret .= $chunk;
 826         }
 827         ++$i;
 828     }
 829     /* remove the first added space */
 830     if ($ret) {
 831         if ($htmlsave) {
 832             $ret = substr($ret,5);
 833         } else {
 834             $ret = substr($ret,1);
 835         }
 836     }
 837
 838     return $ret;
 839 }
 840
 841 /**
 842  * Encodes header
 843  *
 844  * Function uses XTRA_CODE _encodeheader function, if such function exists.
 845  *
 846  * Function uses Q encoding by default and encodes a string according to RFC
 847  * 1522 for use in headers if it contains 8-bit characters or anything that
 848  * looks like it should be encoded.
 849  *
 850  * Function switches to B encoding and encodeHeaderBase64() function, if
 851  * string is 8bit and multibyte character set supported by mbstring extension
 852  * is used. It can cause E_USER_NOTICE errors, if interface is used with
 853  * multibyte character set unsupported by mbstring extension.
 854  *
 855  * @param string $string header string, that has to be encoded
 856  * @return string quoted-printable encoded string
 857  * @todo make $mb_charsets system wide constant
 858  */
 859 function encodeHeader ($string) {
 860     global $default_charset, $languages, $squirrelmail_language;
 861
 862     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 863             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
 864         return  call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
 865     }
 866
 867     // Use B encoding for multibyte charsets
 868     $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
 869     if (in_array($default_charset,$mb_charsets) &&
 870         in_array($default_charset,sq_mb_list_encodings()) &&
 871         sq_is8bit($string)) {
 872         return encodeHeaderBase64($string,$default_charset);
 873     } elseif (in_array($default_charset,$mb_charsets) &&
 874               sq_is8bit($string) &&
 875               ! in_array($default_charset,sq_mb_list_encodings())) {
 876         // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
 877         // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
 878     }
 879
 880     // Encode only if the string contains 8-bit characters or =?
 881     $j = strlen($string);
 882     $max_l = 75 - strlen($default_charset) - 7;
 883     $aRet = array();
 884     $ret = '';
 885     $iEncStart = $enc_init = false;
 886     $cur_l = $iOffset = 0;
 887     for($i = 0; $i < $j; ++$i) {
 888         switch($string{$i})
 889         {
 890             case '=':
 891             case '<':
 892             case '>':
 893             case ',':
 894             case '?':
 895             case '_':
 896                 if ($iEncStart === false) {
 897                     $iEncStart = $i;
 898                 }
 899                 $cur_l+=3;
 900                 if ($cur_l > ($max_l-2)) {
 901                     /* if there is an stringpart that doesn't need encoding, add it */
 902                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 903                     $aRet[] = "=?$default_charset?Q?$ret?=";
 904                     $iOffset = $i;
 905                     $cur_l = 0;
 906                     $ret = '';
 907                     $iEncStart = false;
 908                 } else {
 909                     $ret .= sprintf("=%02X",ord($string{$i}));
 910                 }
 911                 break;
 912             case '(':
 913             case ')':
 914                 if ($iEncStart !== false) {
 915                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 916                     $aRet[] = "=?$default_charset?Q?$ret?=";
 917                     $iOffset = $i;
 918                     $cur_l = 0;
 919                     $ret = '';
 920                     $iEncStart = false;
 921                 }
 922                 break;
 923             case ' ':
 924                 if ($iEncStart !== false) {
 925                     $cur_l++;
 926                     if ($cur_l > $max_l) {
 927                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 928                         $aRet[] = "=?$default_charset?Q?$ret?=";
 929                         $iOffset = $i;
 930                         $cur_l = 0;
 931                         $ret = '';
 932                         $iEncStart = false;
 933                     } else {
 934                         $ret .= '_';
 935                     }
 936                 }
 937                 break;
 938             default:
 939                 $k = ord($string{$i});
 940                 if ($k > 126) {
 941                     if ($iEncStart === false) {
 942                         // do not start encoding in the middle of a string, also take the rest of the word.
 943                         $sLeadString = substr($string,0,$i);
 944                         $aLeadString = explode(' ',$sLeadString);
 945                         $sToBeEncoded = array_pop($aLeadString);
 946                         $iEncStart = $i - strlen($sToBeEncoded);
 947                         $ret .= $sToBeEncoded;
 948                         $cur_l += strlen($sToBeEncoded);
 949                     }
 950                     $cur_l += 3;
 951                     /* first we add the encoded string that reached it's max size */
 952                     if ($cur_l > ($max_l-2)) {
 953                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 954                         $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
 955                         $cur_l = 3;
 956                         $ret = '';
 957                         $iOffset = $i;
 958                         $iEncStart = $i;
 959                     }
 960                     $enc_init = true;
 961                     $ret .= sprintf("=%02X", $k);
 962                 } else {
 963                     if ($iEncStart !== false) {
 964                         $cur_l++;
 965                         if ($cur_l > $max_l) {
 966                             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 967                             $aRet[] = "=?$default_charset?Q?$ret?=";
 968                             $iEncStart = false;
 969                             $iOffset = $i;
 970                             $cur_l = 0;
 971                             $ret = '';
 972                         } else {
 973                             $ret .= $string{$i};
 974                         }
 975                     }
 976                 }
 977                 break;
 978         }
 979     }
 980
 981     if ($enc_init) {
 982         if ($iEncStart !== false) {
 983             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 984             $aRet[] = "=?$default_charset?Q?$ret?=";
 985         } else {
 986             $aRet[] = substr($string,$iOffset);
 987         }
 988         $string = implode('',$aRet);
 989     }
 990     return $string;
 991 }
 992
 993 /**
 994  * Encodes string according to rfc2047 B encoding header formating rules
 995  *
 996  * It is recommended way to encode headers with character sets that store
 997  * symbols in more than one byte.
 998  *
 999  * Function requires mbstring support. If required mbstring functions are missing,
1000  * function returns false and sets E_USER_WARNING level error message.
1001  *
1002  * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
1003  * that mbstring functions will generate E_WARNING errors, if unsupported
1004  * character set is used. mb_encode_mimeheader function provided by php
1005  * mbstring extension is not used in order to get better control of header
1006  * encoding.
1007  *
1008  * Used php code functions - function_exists(), trigger_error(), strlen()
1009  * (is used with charset names and base64 strings). Used php mbstring
1010  * functions - mb_strlen and mb_substr.
1011  *
1012  * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
1013  * encoding), rfc 2822 (header folding)
1014  *
1015  * @param string $string header string that must be encoded
1016  * @param string $charset character set. Must be supported by mbstring extension.
1017  * Use sq_mb_list_encodings() to detect supported charsets.
1018  * @return string string encoded according to rfc2047 B encoding formating rules
1019  * @since 1.5.1
1020  * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1021  * @todo Do we want to control max length of header?
1022  * @todo Do we want to control EOL (end-of-line) marker?
1023  * @todo Do we want to translate error message?
1024  */
1025 function encodeHeaderBase64($string,$charset) {
1026     /**
1027      * Check mbstring function requirements.
1028      */
1029     if (! function_exists('mb_strlen') ||
1030         ! function_exists('mb_substr')) {
1031         // set E_USER_WARNING
1032         trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
1033         // return false
1034         return false;
1035     }
1036
1037     // initial return array
1038     $aRet = array();
1039
1040     /**
1041      * header length = 75 symbols max (same as in encodeHeader)
1042      * remove $charset length
1043      * remove =? ? ?= (5 chars)
1044      * remove 2 more chars (\r\n ?)
1045      */
1046     $iMaxLength = 75 - strlen($charset) - 7;
1047
1048     // set first character position
1049     $iStartCharNum = 0;
1050
1051     // loop through all characters. count characters and not bytes.
1052     for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1053         // encode string from starting character to current character.
1054         $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1055
1056         // Check encoded string length
1057         if(strlen($encoded_string)>$iMaxLength) {
1058             // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1059             $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1060
1061             // set new starting character
1062             $iStartCharNum = $iCharNum-1;
1063
1064             // encode last char (in case it is last character in string)
1065             $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1066         } // if string is shorter than max length - add next character
1067     }
1068
1069     // add last encoded string to array
1070     $aRet[] = $encoded_string;
1071
1072     // set initial return string
1073     $sRet = '';
1074
1075     // loop through encoded strings
1076     foreach($aRet as $string) {
1077         // TODO: Do we want to control EOL (end-of-line) marker
1078         if ($sRet!='') $sRet.= " ";
1079
1080         // add header tags and encoded string to return string
1081         $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1082     }
1083
1084     return $sRet;
1085 }
1086
1087 /* This function trys to locate the entity_id of a specific mime element */
1088 function find_ent_id($id, $message) {
1089     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1090         if ($message->entities[$i]->header->type0 == 'multipart')  {
1091             $ret = find_ent_id($id, $message->entities[$i]);
1092         } else {
1093             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
1094 //                if (sq_check_save_extension($message->entities[$i])) {
1095                 return $message->entities[$i]->entity_id;
1096 //                }
1097             } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1098                 /**
1099                  * This is part of a fix for Outlook Express 6.x generating
1100                  * cid URLs without creating content-id headers
1101                  * @@JA - 20050207
1102                  */
1103                 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1104                     return $message->entities[$i]->entity_id;
1105                 }
1106             }
1107         }
1108     }
1109     return $ret;
1110 }
1111
1112 function sq_check_save_extension($message) {
1113     $filename = $message->getFilename();
1114     $ext = substr($filename, strrpos($filename,'.')+1);
1115     $save_extensions = array('jpg','jpeg','gif','png','bmp');
1116     return in_array($ext, $save_extensions);
1117 }
1118
1119
1120 /**
1121  ** HTMLFILTER ROUTINES
1122  */
1123
1124 /**
1125  * This function checks attribute values for entity-encoded values
1126  * and returns them translated into 8-bit strings so we can run
1127  * checks on them.
1128  *
1129  * @param  $attvalue A string to run entity check against.
1130  * @return           Nothing, modifies a reference value.
1131  */
1132 function sq_defang(&$attvalue){
1133     $me = 'sq_defang';
1134     /**
1135      * Skip this if there aren't ampersands or backslashes.
1136      */
1137     if (strpos($attvalue, '&') === false
1138         && strpos($attvalue, '\\') === false){
1139         return;
1140     }
1141     $m = false;
1142     do {
1143         $m = false;
1144         $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1145         $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1146         $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1147     } while ($m == true);
1148     $attvalue = stripslashes($attvalue);
1149 }
1150
1151 /**
1152  * Kill any tabs, newlines, or carriage returns. Our friends the
1153  * makers of the browser with 95% market value decided that it'd
1154  * be funny to make "java[tab]script" be just as good as "javascript".
1155  *
1156  * @param  attvalue  The attribute value before extraneous spaces removed.
1157  * @return attvalue  Nothing, modifies a reference value.
1158  */
1159 function sq_unspace(&$attvalue){
1160     $me = 'sq_unspace';
1161     if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1162         $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1163                                 Array('',   '',   '',   '',   ''), $attvalue);
1164     }
1165 }
1166
1167 /**
1168  * This function returns the final tag out of the tag name, an array
1169  * of attributes, and the type of the tag. This function is called by
1170  * sq_sanitize internally.
1171  *
1172  * @param  $tagname  the name of the tag.
1173  * @param  $attary   the array of attributes and their values
1174  * @param  $tagtype  The type of the tag (see in comments).
1175  * @return           a string with the final tag representation.
1176  */
1177 function sq_tagprint($tagname, $attary, $tagtype){
1178     $me = 'sq_tagprint';
1179
1180     if ($tagtype == 2){
1181         $fulltag = '</' . $tagname . '>';
1182     } else {
1183         $fulltag = '<' . $tagname;
1184         if (is_array($attary) && sizeof($attary)){
1185             $atts = Array();
1186             while (list($attname, $attvalue) = each($attary)){
1187                 array_push($atts, "$attname=$attvalue");
1188             }
1189             $fulltag .= ' ' . join(" ", $atts);
1190         }
1191         if ($tagtype == 3){
1192             $fulltag .= ' /';
1193         }
1194         $fulltag .= '>';
1195     }
1196     return $fulltag;
1197 }
1198
1199 /**
1200  * A small helper function to use with array_walk. Modifies a by-ref
1201  * value and makes it lowercase.
1202  *
1203  * @param  $val a value passed by-ref.
1204  * @return      void since it modifies a by-ref value.
1205  */
1206 function sq_casenormalize(&$val){
1207     $val = strtolower($val);
1208 }
1209
1210 /**
1211  * This function skips any whitespace from the current position within
1212  * a string and to the next non-whitespace value.
1213  *
1214  * @param  $body   the string
1215  * @param  $offset the offset within the string where we should start
1216  *                 looking for the next non-whitespace character.
1217  * @return         the location within the $body where the next
1218  *                 non-whitespace char is located.
1219  */
1220 function sq_skipspace($body, $offset){
1221     $me = 'sq_skipspace';
1222     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1223     if (sizeof($matches{1})){
1224         $count = strlen($matches{1});
1225         $offset += $count;
1226     }
1227     return $offset;
1228 }
1229
1230 /**
1231  * This function looks for the next character within a string.  It's
1232  * really just a glorified "strpos", except it catches if failures
1233  * nicely.
1234  *
1235  * @param  $body   The string to look for needle in.
1236  * @param  $offset Start looking from this position.
1237  * @param  $needle The character/string to look for.
1238  * @return         location of the next occurance of the needle, or
1239  *                 strlen($body) if needle wasn't found.
1240  */
1241 function sq_findnxstr($body, $offset, $needle){
1242     $me  = 'sq_findnxstr';
1243     $pos = strpos($body, $needle, $offset);
1244     if ($pos === FALSE){
1245         $pos = strlen($body);
1246     }
1247     return $pos;
1248 }
1249
1250 /**
1251  * This function takes a PCRE-style regexp and tries to match it
1252  * within the string.
1253  *
1254  * @param  $body   The string to look for needle in.
1255  * @param  $offset Start looking from here.
1256  * @param  $reg    A PCRE-style regex to match.
1257  * @return         Returns a false if no matches found, or an array
1258  *                 with the following members:
1259  *                 - integer with the location of the match within $body
1260  *                 - string with whatever content between offset and the match
1261  *                 - string with whatever it is we matched
1262  */
1263 function sq_findnxreg($body, $offset, $reg){
1264     $me = 'sq_findnxreg';
1265     $matches = Array();
1266     $retarr = Array();
1267     preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1268     if (!isset($matches{0}) || !$matches{0}){
1269         $retarr = false;
1270     } else {
1271         $retarr{0} = $offset + strlen($matches{1});
1272         $retarr{1} = $matches{1};
1273         $retarr{2} = $matches{2};
1274     }
1275     return $retarr;
1276 }
1277
1278 /**
1279  * This function looks for the next tag.
1280  *
1281  * @param  $body   String where to look for the next tag.
1282  * @param  $offset Start looking from here.
1283  * @return         false if no more tags exist in the body, or
1284  *                 an array with the following members:
1285  *                 - string with the name of the tag
1286  *                 - array with attributes and their values
1287  *                 - integer with tag type (1, 2, or 3)
1288  *                 - integer where the tag starts (starting "<")
1289  *                 - integer where the tag ends (ending ">")
1290  *                 first three members will be false, if the tag is invalid.
1291  */
1292 function sq_getnxtag($body, $offset){
1293     $me = 'sq_getnxtag';
1294     if ($offset > strlen($body)){
1295         return false;
1296     }
1297     $lt = sq_findnxstr($body, $offset, "<");
1298     if ($lt == strlen($body)){
1299         return false;
1300     }
1301     /**
1302      * We are here:
1303      * blah blah <tag attribute="value">
1304      * \---------^
1305      */
1306     $pos = sq_skipspace($body, $lt+1);
1307     if ($pos >= strlen($body)){
1308         return Array(false, false, false, $lt, strlen($body));
1309     }
1310     /**
1311      * There are 3 kinds of tags:
1312      * 1. Opening tag, e.g.:
1313      *    <a href="blah">
1314      * 2. Closing tag, e.g.:
1315      *    </a>
1316      * 3. XHTML-style content-less tag, e.g.:
1317      *    <img src="blah" />
1318      */
1319     $tagtype = false;
1320     switch (substr($body, $pos, 1)){
1321         case '/':
1322             $tagtype = 2;
1323             $pos++;
1324             break;
1325         case '!':
1326             /**
1327              * A comment or an SGML declaration.
1328              */
1329             if (substr($body, $pos+1, 2) == "--"){
1330                 $gt = strpos($body, "-->", $pos);
1331                 if ($gt === false){
1332                     $gt = strlen($body);
1333                 } else {
1334                     $gt += 2;
1335                 }
1336                 return Array(false, false, false, $lt, $gt);
1337             } else {
1338                 $gt = sq_findnxstr($body, $pos, ">");
1339                 return Array(false, false, false, $lt, $gt);
1340             }
1341             break;
1342         default:
1343             /**
1344              * Assume tagtype 1 for now. If it's type 3, we'll switch values
1345              * later.
1346              */
1347             $tagtype = 1;
1348             break;
1349     }
1350
1351     $tag_start = $pos;
1352     $tagname = '';
1353     /**
1354      * Look for next [\W-_], which will indicate the end of the tag name.
1355      */
1356     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1357     if ($regary == false){
1358         return Array(false, false, false, $lt, strlen($body));
1359     }
1360     list($pos, $tagname, $match) = $regary;
1361     $tagname = strtolower($tagname);
1362
1363     /**
1364      * $match can be either of these:
1365      * '>'  indicating the end of the tag entirely.
1366      * '\s' indicating the end of the tag name.
1367      * '/'  indicating that this is type-3 xhtml tag.
1368      *
1369      * Whatever else we find there indicates an invalid tag.
1370      */
1371     switch ($match){
1372         case '/':
1373             /**
1374              * This is an xhtml-style tag with a closing / at the
1375              * end, like so: <img src="blah" />. Check if it's followed
1376              * by the closing bracket. If not, then this tag is invalid
1377              */
1378             if (substr($body, $pos, 2) == "/>"){
1379                 $pos++;
1380                 $tagtype = 3;
1381             } else {
1382                 $gt = sq_findnxstr($body, $pos, ">");
1383                 $retary = Array(false, false, false, $lt, $gt);
1384                 return $retary;
1385             }
1386         case '>':
1387             return Array($tagname, false, $tagtype, $lt, $pos);
1388             break;
1389         default:
1390             /**
1391              * Check if it's whitespace
1392              */
1393             if (!preg_match('/\s/', $match)){
1394                 /**
1395                  * This is an invalid tag! Look for the next closing ">".
1396                  */
1397                 $gt = sq_findnxstr($body, $lt, ">");
1398                 return Array(false, false, false, $lt, $gt);
1399             }
1400             break;
1401     }
1402
1403     /**
1404      * At this point we're here:
1405      * <tagname  attribute='blah'>
1406      * \-------^
1407      *
1408      * At this point we loop in order to find all attributes.
1409      */
1410     $attname = '';
1411     $atttype = false;
1412     $attary = Array();
1413
1414     while ($pos <= strlen($body)){
1415         $pos = sq_skipspace($body, $pos);
1416         if ($pos == strlen($body)){
1417             /**
1418              * Non-closed tag.
1419              */
1420             return Array(false, false, false, $lt, $pos);
1421         }
1422         /**
1423          * See if we arrived at a ">" or "/>", which means that we reached
1424          * the end of the tag.
1425          */
1426         $matches = Array();
1427         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1428             /**
1429              * Yep. So we did.
1430              */
1431             $pos += strlen($matches{1});
1432             if ($matches{2} == "/>"){
1433                 $tagtype = 3;
1434                 $pos++;
1435             }
1436             return Array($tagname, $attary, $tagtype, $lt, $pos);
1437         }
1438
1439         /**
1440          * There are several types of attributes, with optional
1441          * [:space:] between members.
1442          * Type 1:
1443          *   attrname[:space:]=[:space:]'CDATA'
1444          * Type 2:
1445          *   attrname[:space:]=[:space:]"CDATA"
1446          * Type 3:
1447          *   attr[:space:]=[:space:]CDATA
1448          * Type 4:
1449          *   attrname
1450          *
1451          * We leave types 1 and 2 the same, type 3 we check for
1452          * '"' and convert to "&quot" if needed, then wrap in
1453          * double quotes. Type 4 we convert into:
1454          * attrname="yes".
1455          */
1456         $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1457         if ($regary == false){
1458             /**
1459              * Looks like body ended before the end of tag.
1460              */
1461             return Array(false, false, false, $lt, strlen($body));
1462         }
1463         list($pos, $attname, $match) = $regary;
1464         $attname = strtolower($attname);
1465         /**
1466          * We arrived at the end of attribute name. Several things possible
1467          * here:
1468          * '>'  means the end of the tag and this is attribute type 4
1469          * '/'  if followed by '>' means the same thing as above
1470          * '\s' means a lot of things -- look what it's followed by.
1471          *      anything else means the attribute is invalid.
1472          */
1473         switch($match){
1474             case '/':
1475                 /**
1476                  * This is an xhtml-style tag with a closing / at the
1477                  * end, like so: <img src="blah" />. Check if it's followed
1478                  * by the closing bracket. If not, then this tag is invalid
1479                  */
1480                 if (substr($body, $pos, 2) == "/>"){
1481                     $pos++;
1482                     $tagtype = 3;
1483                 } else {
1484                     $gt = sq_findnxstr($body, $pos, ">");
1485                     $retary = Array(false, false, false, $lt, $gt);
1486                     return $retary;
1487                 }
1488             case '>':
1489                 $attary{$attname} = '"yes"';
1490                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1491                 break;
1492             default:
1493                 /**
1494                  * Skip whitespace and see what we arrive at.
1495                  */
1496                 $pos = sq_skipspace($body, $pos);
1497                 $char = substr($body, $pos, 1);
1498                 /**
1499                  * Two things are valid here:
1500                  * '=' means this is attribute type 1 2 or 3.
1501                  * \w means this was attribute type 4.
1502                  * anything else we ignore and re-loop. End of tag and
1503                  * invalid stuff will be caught by our checks at the beginning
1504                  * of the loop.
1505                  */
1506                 if ($char == "="){
1507                     $pos++;
1508                     $pos = sq_skipspace($body, $pos);
1509                     /**
1510                      * Here are 3 possibilities:
1511                      * "'"  attribute type 1
1512                      * '"'  attribute type 2
1513                      * everything else is the content of tag type 3
1514                      */
1515                     $quot = substr($body, $pos, 1);
1516                     if ($quot == "'"){
1517                         $regary = sq_findnxreg($body, $pos+1, "\'");
1518                         if ($regary == false){
1519                             return Array(false, false, false, $lt, strlen($body));
1520                         }
1521                         list($pos, $attval, $match) = $regary;
1522                         $pos++;
1523                         $attary{$attname} = "'" . $attval . "'";
1524                     } else if ($quot == '"'){
1525                         $regary = sq_findnxreg($body, $pos+1, '\"');
1526                         if ($regary == false){
1527                             return Array(false, false, false, $lt, strlen($body));
1528                         }
1529                         list($pos, $attval, $match) = $regary;
1530                         $pos++;
1531                         $attary{$attname} = '"' . $attval . '"';
1532                     } else {
1533                         /**
1534                          * These are hateful. Look for \s, or >.
1535                          */
1536                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1537                         if ($regary == false){
1538                             return Array(false, false, false, $lt, strlen($body));
1539                         }
1540                         list($pos, $attval, $match) = $regary;
1541                         /**
1542                          * If it's ">" it will be caught at the top.
1543                          */
1544                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1545                         $attary{$attname} = '"' . $attval . '"';
1546                     }
1547                 } else if (preg_match("|[\w/>]|", $char)) {
1548                     /**
1549                      * That was attribute type 4.
1550                      */
1551                     $attary{$attname} = '"yes"';
1552                 } else {
1553                     /**
1554                      * An illegal character. Find next '>' and return.
1555                      */
1556                     $gt = sq_findnxstr($body, $pos, ">");
1557                     return Array(false, false, false, $lt, $gt);
1558                 }
1559                 break;
1560         }
1561     }
1562     /**
1563      * The fact that we got here indicates that the tag end was never
1564      * found. Return invalid tag indication so it gets stripped.
1565      */
1566     return Array(false, false, false, $lt, strlen($body));
1567 }
1568
1569 /**
1570  * Translates entities into literal values so they can be checked.
1571  *
1572  * @param $attvalue the by-ref value to check.
1573  * @param $regex    the regular expression to check against.
1574  * @param $hex      whether the entites are hexadecimal.
1575  * @return          True or False depending on whether there were matches.
1576  */
1577 function sq_deent(&$attvalue, $regex, $hex=false){
1578     $me = 'sq_deent';
1579     $ret_match = false;
1580     preg_match_all($regex, $attvalue, $matches);
1581     if (is_array($matches) && sizeof($matches[0]) > 0){
1582         $repl = Array();
1583         for ($i = 0; $i < sizeof($matches[0]); $i++){
1584             $numval = $matches[1][$i];
1585             if ($hex){
1586                 $numval = hexdec($numval);
1587             }
1588             $repl{$matches[0][$i]} = chr($numval);
1589         }
1590         $attvalue = strtr($attvalue, $repl);
1591         return true;
1592     } else {
1593         return false;
1594     }
1595 }
1596
1597 /**
1598  * This function runs various checks against the attributes.
1599  *
1600  * @param  $tagname         String with the name of the tag.
1601  * @param  $attary          Array with all tag attributes.
1602  * @param  $rm_attnames     See description for sq_sanitize
1603  * @param  $bad_attvals     See description for sq_sanitize
1604  * @param  $add_attr_to_tag See description for sq_sanitize
1605  * @param  $message         message object
1606  * @param  $id              message id
1607  * @return                  Array with modified attributes.
1608  */
1609 function sq_fixatts($tagname,
1610                     $attary,
1611                     $rm_attnames,
1612                     $bad_attvals,
1613                     $add_attr_to_tag,
1614                     $message,
1615                     $id,
1616                     $mailbox
1617                     ){
1618     $me = 'sq_fixatts';
1619     while (list($attname, $attvalue) = each($attary)){
1620         /**
1621          * See if this attribute should be removed.
1622          */
1623         foreach ($rm_attnames as $matchtag=>$matchattrs){
1624             if (preg_match($matchtag, $tagname)){
1625                 foreach ($matchattrs as $matchattr){
1626                     if (preg_match($matchattr, $attname)){
1627                         unset($attary{$attname});
1628                         continue;
1629                     }
1630                 }
1631             }
1632         }
1633         /**
1634          * Remove any backslashes, entities, and extraneous whitespace.
1635          */
1636         sq_defang($attvalue);
1637         sq_unspace($attvalue);
1638
1639         /**
1640          * Now let's run checks on the attvalues.
1641          * I don't expect anyone to comprehend this. If you do,
1642          * get in touch with me so I can drive to where you live and
1643          * shake your hand personally. :)
1644          */
1645         foreach ($bad_attvals as $matchtag=>$matchattrs){
1646             if (preg_match($matchtag, $tagname)){
1647                 foreach ($matchattrs as $matchattr=>$valary){
1648                     if (preg_match($matchattr, $attname)){
1649                         /**
1650                          * There are two arrays in valary.
1651                          * First is matches.
1652                          * Second one is replacements
1653                          */
1654                         list($valmatch, $valrepl) = $valary;
1655                         $newvalue =
1656                             preg_replace($valmatch, $valrepl, $attvalue);
1657                         if ($newvalue != $attvalue){
1658                             $attary{$attname} = $newvalue;
1659                         }
1660                     }
1661                 }
1662             }
1663         }
1664
1665         /**
1666          * Replace empty src tags with the blank image.  src is only used
1667          * for frames, images, and image inputs.  Doing a replace should
1668          * not affect them working as should be, however it will stop
1669          * IE from being kicked off when src for img tags are not set
1670          */
1671         if (($attname == 'src') && ($attvalue == '""')) {
1672             $attary{$attname} = '"' . SM_PATH . 'images/blank.png"';
1673         }
1674
1675         /**
1676          * Turn cid: urls into http-friendly ones.
1677          */
1678         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1679             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1680         }
1681
1682         /**
1683          * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1684          * One day MS might actually make it match something useful, for now, falling
1685          * back to using cid2http, so we can grab the blank.png.
1686          */
1687         if (preg_match("/^[\'\"]\s*outbind:\/\//si", $attvalue)) {
1688             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1689         }
1690
1691     }
1692     /**
1693      * See if we need to append any attributes to this tag.
1694      */
1695     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1696         if (preg_match($matchtag, $tagname)){
1697             $attary = array_merge($attary, $addattary);
1698         }
1699     }
1700     return $attary;
1701 }
1702
1703 /**
1704  * This function edits the style definition to make them friendly and
1705  * usable in SquirrelMail.
1706  *
1707  * @param  $message  the message object
1708  * @param  $id       the message id
1709  * @param  $content  a string with whatever is between <style> and </style>
1710  * @param  $mailbox  the message mailbox
1711  * @return           a string with edited content.
1712  */
1713 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1714     global $view_unsafe_images;
1715     $me = 'sq_fixstyle';
1716     $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1717     if ($ret == FALSE){
1718         return array(FALSE, strlen($body));
1719     }
1720     $newpos = $ret[0] + strlen($ret[2]);
1721     $content = $ret[1];
1722     /**
1723     * First look for general BODY style declaration, which would be
1724     * like so:
1725     * body {background: blah-blah}
1726     * and change it to .bodyclass so we can just assign it to a <div>
1727     */
1728     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1729     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1730     /**
1731     * Fix url('blah') declarations.
1732     */
1733     //   $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1734     //                           "url(\\1$secremoveimg\\2)", $content);
1735     // remove NUL
1736     $content = str_replace("\0", "", $content);
1737     // translate ur\l and variations (IE parses that)
1738     $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
1739     // NB I insert NUL characters to keep to avoid an infinite loop. They are removed after the loop.
1740     while (preg_match("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si", $content, $matches)) {
1741         $sProto = strtolower($matches[1]);
1742         switch ($sProto) {
1743             /**
1744              * Fix url('https*://.*) declarations but only if $view_unsafe_images
1745              * is false.
1746              */
1747             case 'https':
1748             case 'http':
1749                 if (!$view_unsafe_images){
1750
1751                     $sExpr = "/url\s*\(\s*[\'\"]?\s*$sProto*:.*[\'\"]?\s*\)/si";
1752                     $content = preg_replace($sExpr, "u\0r\0l(\\1$secremoveimg\\2)", $content);
1753
1754                 } else {
1755                     $content = preg_replace('/url/i',"u\0r\0l",$content);
1756                 }
1757                 break;
1758             /**
1759              * Fix urls that refer to cid:
1760              */
1761             case 'cid':
1762                 $cidurl = 'cid:'. $matches[2];
1763                 $httpurl = sq_cid2http($message, $id, $cidurl, $mailbox);
1764                 // escape parentheses that can modify the regular expression
1765                 $cidurl = str_replace(array('(',')'),array('\\(','\\)'),$cidurl);
1766                 $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1767                                         "u\0r\0l($httpurl)", $content);
1768                 break;
1769             default:
1770                 /**
1771                  * replace url with protocol other then the white list
1772                  * http,https and cid by an empty string.
1773                  */
1774                 $content = preg_replace("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si",
1775                                  "", $content);
1776                 break;
1777         }
1778     }
1779     // remove NUL
1780     $content = str_replace("\0", "", $content);
1781    /**
1782     * Remove any backslashes, entities, and extraneous whitespace.
1783     */
1784     $contentTemp = $content;
1785     sq_defang($contentTemp);
1786     sq_unspace($contentTemp);
1787
1788     /**
1789      * Fix stupid css declarations which lead to vulnerabilities
1790      * in IE.
1791      */
1792     $match   = Array('/\/\*.*\*\//',
1793                     '/expression/i',
1794                     '/behaviou*r/i',
1795                     '/binding/i',
1796                     '/include-source/i');
1797     $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy');
1798     $contentNew = preg_replace($match, $replace, $contentTemp);
1799     if ($contentNew !== $contentTemp) {
1800         // insecure css declarations are used. From now on we don't care
1801         // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
1802         $content = $contentNew;
1803     }
1804     return array($content, $newpos);
1805 }
1806
1807
1808 /**
1809  * This function converts cid: url's into the ones that can be viewed in
1810  * the browser.
1811  *
1812  * @param  $message  the message object
1813  * @param  $id       the message id
1814  * @param  $cidurl   the cid: url.
1815  * @param  $mailbox  the message mailbox
1816  * @return           a string with a http-friendly url
1817  */
1818 function sq_cid2http($message, $id, $cidurl, $mailbox){
1819     /**
1820      * Get rid of quotes.
1821      */
1822     $quotchar = substr($cidurl, 0, 1);
1823     if ($quotchar == '"' || $quotchar == "'"){
1824         $cidurl = str_replace($quotchar, "", $cidurl);
1825     } else {
1826         $quotchar = '';
1827     }
1828     $cidurl = substr(trim($cidurl), 4);
1829
1830     $match_str = '/\{.*?\}\//';
1831     $str_rep = '';
1832     $cidurl = preg_replace($match_str, $str_rep, $cidurl);
1833
1834     $linkurl = find_ent_id($cidurl, $message);
1835     /* in case of non-save cid links $httpurl should be replaced by a sort of
1836        unsave link image */
1837     $httpurl = '';
1838
1839     /**
1840      * This is part of a fix for Outlook Express 6.x generating
1841      * cid URLs without creating content-id headers. These images are
1842      * not part of the multipart/related html mail. The html contains
1843      * <img src="cid:{some_id}/image_filename.ext"> references to
1844      * attached images with as goal to render them inline although
1845      * the attachment disposition property is not inline.
1846      */
1847
1848     if (empty($linkurl)) {
1849         if (preg_match('/{.*}\//', $cidurl)) {
1850             $cidurl = preg_replace('/{.*}\//','', $cidurl);
1851             if (!empty($cidurl)) {
1852                 $linkurl = find_ent_id($cidurl, $message);
1853             }
1854         }
1855     }
1856
1857     if (!empty($linkurl)) {
1858         $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1859             "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1860             '&amp;ent_id=' . $linkurl . $quotchar;
1861     } else {
1862         /**
1863          * If we couldn't generate a proper img url, drop in a blank image
1864          * instead of sending back empty, otherwise it causes unusual behaviour
1865          */
1866         $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
1867     }
1868
1869     return $httpurl;
1870 }
1871
1872 /**
1873  * This function changes the <body> tag into a <div> tag since we
1874  * can't really have a body-within-body.
1875  *
1876  * @param  $attary   an array of attributes and values of <body>
1877  * @param  $mailbox  mailbox we're currently reading (for cid2http)
1878  * @param  $message  current message (for cid2http)
1879  * @param  $id       current message id (for cid2http)
1880  * @return           a modified array of attributes to be set for <div>
1881  */
1882 function sq_body2div($attary, $mailbox, $message, $id){
1883     $me = 'sq_body2div';
1884     $divattary = Array('class' => "'bodyclass'");
1885     $text = '#000000';
1886     $has_bgc_stl = $has_txt_stl = false;
1887     $styledef = '';
1888     if (is_array($attary) && sizeof($attary) > 0){
1889         foreach ($attary as $attname=>$attvalue){
1890             $quotchar = substr($attvalue, 0, 1);
1891             $attvalue = str_replace($quotchar, "", $attvalue);
1892             switch ($attname){
1893                 case 'background':
1894                     $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1895                     $styledef .= "background-image: url('$attvalue'); ";
1896                     break;
1897                 case 'bgcolor':
1898                     $has_bgc_stl = true;
1899                     $styledef .= "background-color: $attvalue; ";
1900                     break;
1901                 case 'text':
1902                     $has_txt_stl = true;
1903                     $styledef .= "color: $attvalue; ";
1904                     break;
1905             }
1906         }
1907         // Outlook defines a white bgcolor and no text color. This can lead to
1908         // white text on a white bg with certain themes.
1909         if ($has_bgc_stl && !$has_txt_stl) {
1910             $styledef .= "color: $text; ";
1911         }
1912         if (strlen($styledef) > 0){
1913             $divattary{"style"} = "\"$styledef\"";
1914         }
1915     }
1916     return $divattary;
1917 }
1918
1919 /**
1920  * This is the main function and the one you should actually be calling.
1921  * There are several variables you should be aware of an which need
1922  * special description.
1923  *
1924  * Since the description is quite lengthy, see it here:
1925  * http://linux.duke.edu/projects/mini/htmlfilter/
1926  *
1927  * @param $body                 the string with HTML you wish to filter
1928  * @param $tag_list             see description above
1929  * @param $rm_tags_with_content see description above
1930  * @param $self_closing_tags    see description above
1931  * @param $force_tag_closing    see description above
1932  * @param $rm_attnames          see description above
1933  * @param $bad_attvals          see description above
1934  * @param $add_attr_to_tag      see description above
1935  * @param $message              message object
1936  * @param $id                   message id
1937  * @return                      sanitized html safe to show on your pages.
1938  */
1939 function sq_sanitize($body,
1940                      $tag_list,
1941                      $rm_tags_with_content,
1942                      $self_closing_tags,
1943                      $force_tag_closing,
1944                      $rm_attnames,
1945                      $bad_attvals,
1946                      $add_attr_to_tag,
1947                      $message,
1948                      $id,
1949                      $mailbox
1950                      ){
1951     $me = 'sq_sanitize';
1952     $rm_tags = array_shift($tag_list);
1953     /**
1954      * Normalize rm_tags and rm_tags_with_content.
1955      */
1956     @array_walk($tag_list, 'sq_casenormalize');
1957     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1958     @array_walk($self_closing_tags, 'sq_casenormalize');
1959     /**
1960      * See if tag_list is of tags to remove or tags to allow.
1961      * false  means remove these tags
1962      * true   means allow these tags
1963      */
1964     $curpos = 0;
1965     $open_tags = Array();
1966     $trusted = "\n<!-- begin sanitized html -->\n";
1967     $skip_content = false;
1968     /**
1969      * Take care of netscape's stupid javascript entities like
1970      * &{alert('boo')};
1971      */
1972     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1973
1974     while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1975         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1976         $free_content = substr($body, $curpos, $lt-$curpos);
1977         /**
1978          * Take care of <style>
1979          */
1980         if ($tagname == "style" && $tagtype == 1){
1981             list($free_content, $curpos) =
1982                 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
1983             if ($free_content != FALSE){
1984                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1985                 $trusted .= $free_content;
1986                 $trusted .= sq_tagprint($tagname, false, 2);
1987             }
1988             continue;
1989         }
1990         if ($skip_content == false){
1991             $trusted .= $free_content;
1992         }
1993         if ($tagname != FALSE){
1994             if ($tagtype == 2){
1995                 if ($skip_content == $tagname){
1996                     /**
1997                      * Got to the end of tag we needed to remove.
1998                      */
1999                     $tagname = false;
2000                     $skip_content = false;
2001                 } else {
2002                     if ($skip_content == false){
2003                         if ($tagname == "body"){
2004                             $tagname = "div";
2005                         }
2006                         if (isset($open_tags{$tagname}) &&
2007                                 $open_tags{$tagname} > 0){
2008                             $open_tags{$tagname}--;
2009                         } else {
2010                             $tagname = false;
2011                         }
2012                     }
2013                 }
2014             } else {
2015                 /**
2016                  * $rm_tags_with_content
2017                  */
2018                 if ($skip_content == false){
2019                     /**
2020                      * See if this is a self-closing type and change
2021                      * tagtype appropriately.
2022                      */
2023                     if ($tagtype == 1
2024                             && in_array($tagname, $self_closing_tags)){
2025                         $tagtype = 3;
2026                     }
2027                     /**
2028                      * See if we should skip this tag and any content
2029                      * inside it.
2030                      */
2031                     if ($tagtype == 1 &&
2032                             in_array($tagname, $rm_tags_with_content)){
2033                         $skip_content = $tagname;
2034                     } else {
2035                         if (($rm_tags == false
2036                                     && in_array($tagname, $tag_list)) ||
2037                                 ($rm_tags == true &&
2038                                  !in_array($tagname, $tag_list))){
2039                             $tagname = false;
2040                         } else {
2041                             /**
2042                              * Convert body into div.
2043                              */
2044                             if ($tagname == "body"){
2045                                 $tagname = "div";
2046                                 $attary = sq_body2div($attary, $mailbox,
2047                                         $message, $id);
2048                             }
2049                             if ($tagtype == 1){
2050                                 if (isset($open_tags{$tagname})){
2051                                     $open_tags{$tagname}++;
2052                                 } else {
2053                                     $open_tags{$tagname}=1;
2054                                 }
2055                             }
2056                             /**
2057                              * This is where we run other checks.
2058                              */
2059                             if (is_array($attary) && sizeof($attary) > 0){
2060                                 $attary = sq_fixatts($tagname,
2061                                                      $attary,
2062                                                      $rm_attnames,
2063                                                      $bad_attvals,
2064                                                      $add_attr_to_tag,
2065                                                      $message,
2066                                                      $id,
2067                                                      $mailbox
2068                                                      );
2069                             }
2070                         }
2071                     }
2072                 }
2073             }
2074             if ($tagname != false && $skip_content == false){
2075                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2076             }
2077         }
2078         $curpos = $gt+1;
2079     }
2080     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2081     if ($force_tag_closing == true){
2082         foreach ($open_tags as $tagname=>$opentimes){
2083             while ($opentimes > 0){
2084                 $trusted .= '</' . $tagname . '>';
2085                 $opentimes--;
2086             }
2087         }
2088         $trusted .= "\n";
2089     }
2090     $trusted .= "<!-- end sanitized html -->\n";
2091     return $trusted;
2092 }
2093
2094 /**
2095  * This is a wrapper function to call html sanitizing routines.
2096  *
2097  * @param  $body  the body of the message
2098  * @param  $id    the id of the message
2099
2100  * @param  $message
2101  * @param  $mailbox
2102  * @param  boolean $take_mailto_links When TRUE, converts mailto: links
2103  *                                    into internal SM compose links
2104  *                                    (optional; default = TRUE)
2105  * @return        a string with html safe to display in the browser.
2106  */
2107 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2108
2109     // require_once(SM_PATH . 'functions/url_parser.php');  // for $MailTo_PReg_Match
2110
2111     global $attachment_common_show_images, $view_unsafe_images,
2112            $has_unsafe_images;
2113     /**
2114      * Don't display attached images in HTML mode.
2115      *
2116      * SB: why?
2117      */
2118     $attachment_common_show_images = false;
2119     $tag_list = Array(
2120             false,
2121             "object",
2122             "meta",
2123             "html",
2124             "head",
2125             "base",
2126             "link",
2127             "frame",
2128             "iframe",
2129             "plaintext",
2130             "marquee"
2131             );
2132
2133     $rm_tags_with_content = Array(
2134             "script",
2135             "applet",
2136             "embed",
2137             "title",
2138             "frameset",
2139             "xmp",
2140             "xml"
2141             );
2142
2143     $self_closing_tags =  Array(
2144             "img",
2145             "br",
2146             "hr",
2147             "input",
2148             "outbind"
2149             );
2150
2151     $force_tag_closing = true;
2152
2153     $rm_attnames = Array(
2154             "/.*/" =>
2155             Array(
2156                 "/target/i",
2157                 "/^on.*/i",
2158                 "/^dynsrc/i",
2159                 "/^data.*/i",
2160                 "/^lowsrc.*/i"
2161                 )
2162             );
2163
2164     $secremoveimg = "../images/" . _("sec_remove_eng.png");
2165     $bad_attvals = Array(
2166             "/.*/" =>
2167             Array(
2168                 "/^src|background/i" =>
2169                 Array(
2170                     Array(
2171                         "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2172                         "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2173                         "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2174                         ),
2175                     Array(
2176                         "\\1$secremoveimg\\2",
2177                         "\\1$secremoveimg\\2",
2178                         "\\1$secremoveimg\\2",
2179                         )
2180                     ),
2181                 "/^href|action/i" =>
2182                 Array(
2183                     Array(
2184                         "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2185                         "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2186                         "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2187                         ),
2188                     Array(
2189                         "\\1#\\1",
2190                         "\\1#\\1",
2191                         "\\1#\\1"
2192                         )
2193                     ),
2194         "/^style/i" =>
2195             Array(
2196                 Array(
2197                     "/\/\*.*\*\//",
2198                     "/expression/i",
2199                     "/binding/i",
2200                     "/behaviou*r/i",
2201                     "/include-source/i",
2202                     "/position\s*:\s*absolute/i",
2203                     "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
2204                     "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2205                     "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2206                     "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2207                     "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2208                     ),
2209                 Array(
2210                     "",
2211                     "idiocy",
2212                     "idiocy",
2213                     "idiocy",
2214                     "idiocy",
2215                     "",
2216                     "url",
2217                     "url(\\1#\\1)",
2218                     "url(\\1#\\1)",
2219                     "url(\\1#\\1)",
2220                     "\\1:url(\\2#\\3)"
2221                     )
2222                 )
2223             )
2224         );
2225     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
2226         $view_unsafe_images = false;
2227     }
2228     if (!$view_unsafe_images){
2229         /**
2230          * Remove any references to http/https if view_unsafe_images set
2231          * to false.
2232          */
2233         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2234                 '/^([\'\"])\s*https*:.*([\'\"])/si');
2235         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2236                 "\\1$secremoveimg\\1");
2237         array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2238                 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2239         array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2240                 "url(\\1$secremoveimg\\1)");
2241     }
2242
2243     $add_attr_to_tag = Array(
2244             "/^a$/i" =>
2245             Array('target'=>'"_blank"',
2246                 'title'=>'"'._("This external link will open in a new window").'"'
2247                 )
2248             );
2249     $trusted = sq_sanitize($body,
2250                            $tag_list,
2251                            $rm_tags_with_content,
2252                            $self_closing_tags,
2253                            $force_tag_closing,
2254                            $rm_attnames,
2255                            $bad_attvals,
2256                            $add_attr_to_tag,
2257                            $message,
2258                            $id,
2259                            $mailbox
2260                            );
2261     if (preg_match("|$secremoveimg|i", $trusted)){
2262         $has_unsafe_images = true;
2263     }
2264
2265     // we want to parse mailto's in HTML output, change to SM compose links
2266     // this is a modified version of code from url_parser.php... but Marc is
2267     // right: we need a better filtering implementation; adding this randomly
2268     // here is not a great solution
2269     //
2270     if ($take_mailto_links) {
2271         // parseUrl($trusted);   // this even parses URLs inside of tags... too aggressive
2272         global $MailTo_PReg_Match;
2273         $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1) ;
2274         if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2275             foreach ($regs[0] as $i => $mailto_before) {
2276                 $mailto_params = $regs[10][$i];
2277                 // get rid of any tailing quote since we have to add send_to to the end
2278                 //
2279                 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2280                     $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2281                 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2282                     $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2283
2284                 if ($regs[1][$i]) {    //if there is an email addr before '?', we need to merge it with the params
2285                     $to = 'to=' . $regs[1][$i];
2286                     if (strpos($mailto_params, 'to=') > -1)    //already a 'to='
2287                         $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2288                     else {
2289                         if ($mailto_params)    //already some params, append to them
2290                             $mailto_params .= '&amp;' . $to;
2291                         else
2292                             $mailto_params .= '?' . $to;
2293                     }
2294                 }
2295
2296                 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2297
2298                 // we'll already have target=_blank, no need to allow comp_in_new
2299                 // here (which would be a lot more work anyway)
2300                 //
2301                 global $compose_new_win;
2302                 $temp_comp_in_new = $compose_new_win;
2303                 $compose_new_win = 0;
2304                 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2305                 $compose_new_win = $temp_comp_in_new;
2306
2307                 // remove <a href=" and anything after the next quote (we only
2308                 // need the uri, not the link HTML) in compose uri
2309                 //
2310                 $comp_uri = substr($comp_uri, 9);
2311                 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2312                 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2313             }
2314         }
2315     }
2316
2317     return $trusted;
2318 }
2319
2320 /**
2321  * function SendDownloadHeaders - send file to the browser
2322  *
2323  * Original Source: SM core src/download.php
2324  * moved here to make it available to other code, and separate
2325  * front end from back end functionality.
2326  *
2327  * @param string $type0 first half of mime type
2328  * @param string $type1 second half of mime type
2329  * @param string $filename filename to tell the browser for downloaded file
2330  * @param boolean $force whether to force the download dialog to pop
2331  * @param optional integer $filesize send the Content-Header and length to the browser
2332  * @return void
2333  */
2334 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2335     global $languages, $squirrelmail_language;
2336     $isIE = $isIE6plus = false;
2337
2338     sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2339
2340     if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2341             strstr($HTTP_USER_AGENT, 'Opera') === false) {
2342         $isIE = true;
2343     }
2344
2345     if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2346         ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2347         $isIE6plus = true;
2348     }
2349
2350     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2351             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2352         $filename =
2353             call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2354     } else {
2355         $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
2356     }
2357
2358     // A Pox on Microsoft and it's Internet Explorer!
2359     //
2360     // IE has lots of bugs with file downloads.
2361     // It also has problems with SSL.  Both of these cause problems
2362     // for us in this function.
2363     //
2364     // See this article on Cache Control headers and SSL
2365     // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2366     //
2367     // The best thing you can do for IE is to upgrade to the latest
2368     // version
2369     //set all the Cache Control Headers for IE
2370     if ($isIE) {
2371         $filename=rawurlencode($filename);
2372         header ("Pragma: public");
2373         header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2374         header ("Cache-Control: post-check=0, pre-check=0", false);
2375         header ("Cache-Control: private");
2376
2377         //set the inline header for IE, we'll add the attachment header later if we need it
2378         header ("Content-Disposition: inline; filename=$filename");
2379     }
2380
2381     if (!$force) {
2382         // Try to show in browser window
2383         header ("Content-Disposition: inline; filename=\"$filename\"");
2384         header ("Content-Type: $type0/$type1; name=\"$filename\"");
2385     } else {
2386         // Try to pop up the "save as" box
2387
2388         // IE makes this hard.  It pops up 2 save boxes, or none.
2389         // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2390         // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2391         // But, according to Microsoft, it is "RFC compliant but doesn't
2392         // take into account some deviations that allowed within the
2393         // specification."  Doesn't that mean RFC non-compliant?
2394         // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2395
2396         // all browsers need the application/octet-stream header for this
2397         header ("Content-Type: application/octet-stream; name=\"$filename\"");
2398
2399         // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2400         // Do not have quotes around filename, but that applied to
2401         // "attachment"... does it apply to inline too?
2402         header ("Content-Disposition: attachment; filename=\"$filename\"");
2403
2404         if ($isIE && !$isIE6plus) {
2405             // This combination seems to work mostly.  IE 5.5 SP 1 has
2406             // known issues (see the Microsoft Knowledge Base)
2407
2408             // This works for most types, but doesn't work with Word files
2409             header ("Content-Type: application/download; name=\"$filename\"");
2410
2411             // These are spares, just in case.  :-)
2412             //header("Content-Type: $type0/$type1; name=\"$filename\"");
2413             //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2414             //header("Content-Type: application/octet-stream; name=\"$filename\"");
2415         } else {
2416             // another application/octet-stream forces download for Netscape
2417             header ("Content-Type: application/octet-stream; name=\"$filename\"");
2418         }
2419     }
2420
2421     //send the content-length header if the calling function provides it
2422     if ($filesize > 0) {
2423         header("Content-Length: $filesize");
2424     }
2425
2426 }  // end fn SendDownloadHeaders