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             if (!isset($color[13])) {
 317                 $color[13] = '#800000';
 318             }
 319             $line = '<font color="' . $color[13] . '">' . $line . '</font>';
 320         } elseif ($quotes) {
 321             if (!isset($color[14])) {
 322                 $color[14] = '#FF0000';
 323             }
 324             $line = '<font color="' . $color[14] . '">' . $line . '</font>';
 325         }
 326
 327         $body_ary[$i] = $line;
 328     }
 329     $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
 330 }
 331
 332 /**
 333  * This returns a parsed string called $body. That string can then
 334  * be displayed as the actual message in the HTML. It contains
 335  * everything needed, including HTML Tags, Attachments at the
 336  * bottom, etc.
 337  *
 338  * Since 1.2.0 function uses message_body hook.
 339  * Till 1.3.0 function included output of formatAttachments().
 340  *
 341  * @param resource $imap_stream imap connection resource
 342  * @param object $message squirrelmail message object
 343  * @param array $color squirrelmail color theme array
 344  * @param integer $wrap_at number of characters per line
 345  * @param string $ent_num (since 1.3.0) message part id
 346  * @param integer $id (since 1.3.0) message id
 347  * @param string $mailbox (since 1.3.0) imap folder name
 348  * @param boolean $clean (since 1.5.1) Do not output stuff that's irrelevant for the printable version.
 349  * @return string html formated message text
 350  */
 351 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX', $clean=FALSE) {
 352     /* This if statement checks for the entity to show as the
 353      * primary message. To add more of them, just put them in the
 354      * order that is their priority.
 355      */
 356     global $startMessage, $languages, $squirrelmail_language,
 357            $show_html_default, $sort, $has_unsafe_images, $passed_ent_id,
 358            $use_iframe, $iframe_height, $download_and_unsafe_link;
 359
 360     // workaround for not updated config.php
 361     if (! isset($use_iframe)) $use_iframe = false;
 362
 363     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
 364         $view_unsafe_images = false;
 365     }
 366
 367     $body = '';
 368     $urlmailbox = urlencode($mailbox);
 369     $body_message = getEntity($message, $ent_num);
 370     if (($body_message->header->type0 == 'text') ||
 371             ($body_message->header->type0 == 'rfc822')) {
 372         $body = mime_fetch_body ($imap_stream, $id, $ent_num);
 373         $body = decodeBody($body, $body_message->header->encoding);
 374
 375         if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 376                 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
 377             if (mb_detect_encoding($body) != 'ASCII') {
 378                 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
 379             }
 380         }
 381         $hookResults = do_hook("message_body", $body);
 382         $body = $hookResults[1];
 383
 384         /* If there are other types that shouldn't be formatted, add
 385          * them here.
 386          */
 387
 388         if ($body_message->header->type1 == 'html') {
 389             if ($show_html_default <> 1) {
 390                 $entity_conv = array('&nbsp;' => ' ',
 391                                      '<p>'    => "\n",
 392                                      '<P>'    => "\n",
 393                                      '<br>'   => "\n",
 394                                      '<BR>'   => "\n",
 395                                      '<br />' => "\n",
 396                                      '<BR />' => "\n",
 397                                      '&gt;'   => '>',
 398                                      '&lt;'   => '<');
 399                 $body = strtr($body, $entity_conv);
 400                 $body = strip_tags($body);
 401                 $body = trim($body);
 402                 translateText($body, $wrap_at,
 403                         $body_message->header->getParameter('charset'));
 404             } elseif ($use_iframe && ! $clean) {
 405                 // $clean is used to remove iframe in printable view.
 406
 407                 /**
 408                  * If we don't add html message between iframe tags,
 409                  * we must detect unsafe images and modify $has_unsafe_images.
 410                  */
 411                 $html_body = magicHTML($body, $id, $message, $mailbox);
 412                 // Convert character set in order to display html mails in different character set
 413                 $html_body = charset_decode($body_message->header->getParameter('charset'),$html_body,false,true);
 414
 415                 // creating iframe url
 416                 $iframeurl=sqm_baseuri().'src/view_html.php?'
 417                     . 'mailbox=' . $urlmailbox
 418                     . '&amp;passed_id=' . $id
 419                     . '&amp;ent_id=' . $ent_num
 420                     . '&amp;view_unsafe_images=' . (int) $view_unsafe_images;
 421
 422                 // adding warning message
 423                 $body = html_tag('div',_("Viewing HTML formatted email"),'center');
 424
 425                 /**
 426                  * height can't be set to 100%, because it does not work as expected when
 427                  * iframe is inside the table. Browsers do not create full height objects
 428                  * even when iframe is not nested. Maybe there is some way to get full size
 429                  * with CSS. Tested in firefox 1.02 and opera 7.53
 430                  *
 431                  * width="100%" does not work as expected, when table width is not set (automatic)
 432                  *
 433                  * tokul: I think <iframe> are safer sandbox than <object>. Objects might
 434                  * need special handling for IE and IE6SP2.
 435                  */
 436                 $body.= "<div><iframe name=\"message_frame\" width=\"100%\" height=\"$iframe_height\" src=\"$iframeurl\""
 437                     .' frameborder="1" marginwidth="0" marginheight="0" scrolling="auto">' . "\n";
 438
 439                 // Message for browsers without iframe support
 440                 //$body.= _("Your browser does not support inline frames.
 441                 // You can view HTML formated message by following below link.");
 442                 //$body.= "<br /><a href=\"$iframeurl\">"._("View HTML Message")."</a>";
 443
 444                 // if browser can't render iframe, it renders html message.
 445                 $body.= $html_body;
 446
 447                 // close iframe
 448                 $body.="</iframe></div>\n";
 449             } else {
 450                 // old way of html rendering
 451                 $body = magicHTML($body, $id, $message, $mailbox);
 452                 /**
 453                  * convert character set. charset_decode does not remove html special chars
 454                  * applied by magicHTML functions and does not sanitize them second time if
 455                  * fourth argument is true.
 456                  */
 457                 $body = charset_decode($body_message->header->getParameter('charset'),$body,false,true);
 458             }
 459         } else {
 460             translateText($body, $wrap_at,
 461                     $body_message->header->getParameter('charset'));
 462         }
 463
 464         // if this is the clean display (i.e. printer friendly), stop here.
 465         if ( $clean ) {
 466             return $body;
 467         }
 468
 469         $download_and_unsafe_link = '';
 470
 471         $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
 472             '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
 473             '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 474         if (isset($passed_ent_id)) {
 475             $link .= '&amp;passed_ent_id='.$passed_ent_id;
 476         }
 477         $download_and_unsafe_link .= '&nbsp;|&nbsp;<a href="download.php?absolute_dl=true&amp;' .
 478             $link . '">' . _("Download this as a file") .  '</a>';
 479         if ($view_unsafe_images) {
 480             $text = _("Hide Unsafe Images");
 481         } else {
 482             if (isset($has_unsafe_images) && $has_unsafe_images) {
 483                 $link .= '&amp;view_unsafe_images=1';
 484                 $text = _("View Unsafe Images");
 485             } else {
 486                 $text = '';
 487             }
 488         }
 489         if($text != '') {
 490             $download_and_unsafe_link .= '&nbsp;|&nbsp;<a href="read_body.php?' . $link . '">' . $text . '</a>';
 491         }
 492     }
 493     return $body;
 494 }
 495
 496 /**
 497  * Displays attachment links and information
 498  *
 499  * Since 1.3.0 function is not included in formatBody() call.
 500  *
 501  * Since 1.0.2 uses attachment $type0/$type1 hook.
 502  * Since 1.2.5 uses attachment $type0/* hook.
 503  * Since 1.5.0 uses attachments_bottom hook.
 504  *
 505  * @param object $message SquirrelMail message object
 506  * @param array $exclude_id message parts that are not attachments.
 507  * @param string $mailbox mailbox name
 508  * @param integer $id message id
 509  * @return string html formated attachment information.
 510  */
 511 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 512     global $where, $what, $startMessage, $color, $passed_ent_id, $base_uri;
 513
 514     $att_ar = $message->getAttachments($exclude_id);
 515
 516     if (!count($att_ar)) return '';
 517
 518     $attachments = '';
 519
 520     $urlMailbox = urlencode($mailbox);
 521
 522     foreach ($att_ar as $att) {
 523         $ent = $att->entity_id;
 524         $header = $att->header;
 525         $type0 = strtolower($header->type0);
 526         $type1 = strtolower($header->type1);
 527         $name = '';
 528         $links['download link']['text'] = _("Download");
 529         $links['download link']['href'] = $base_uri .
 530             "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 531         if ($type0 =='message' && $type1 == 'rfc822') {
 532             $default_page = $base_uri  . 'src/read_body.php';
 533             $rfc822_header = $att->rfc822_header;
 534             $filename = $rfc822_header->subject;
 535             if (trim( $filename ) == '') {
 536                 $filename = 'untitled-[' . $ent . ']' ;
 537             }
 538             $from_o = $rfc822_header->from;
 539             if (is_object($from_o)) {
 540                 $from_name = decodeHeader($from_o->getAddress(false));
 541             } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
 542                 // something weird happens when a digest message is opened and you return to the digest
 543                 // now the from object is part of an array. Probably the parseHeader call overwrites the info
 544                 // retrieved from the bodystructure in a different way. We need to fix this later.
 545                 // possible starting point, do not fetch header we already have and inspect how
 546                 // the rfc822_header object behaves.
 547                 $from_name = decodeHeader($from_o[0]->getAddress(false));
 548             } else {
 549                 $from_name = _("Unknown sender");
 550             }
 551             $description = $from_name;
 552         } else {
 553             $default_page = $base_uri  . 'src/download.php';
 554             $filename = $att->getFilename();
 555             if ($header->description) {
 556                 $description = decodeHeader($header->description);
 557             } else {
 558                 $description = '';
 559             }
 560         }
 561
 562         $display_filename = $filename;
 563         if (isset($passed_ent_id)) {
 564             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 565         } else {
 566             $passed_ent_id_link = '';
 567         }
 568         $defaultlink = $default_page . "?startMessage=$startMessage"
 569             . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 570             . '&amp;ent_id='.$ent.$passed_ent_id_link;
 571         if ($where && $what) {
 572             $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 573         }
 574
 575         /* This executes the attachment hook with a specific MIME-type.
 576          * If that doesn't have results, it tries if there's a rule
 577          * for a more generic type.
 578          */
 579         $hookresults = do_hook("attachment $type0/$type1", $links,
 580                 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 581                 $display_filename, $where, $what);
 582         if(count($hookresults[1]) <= 1) {
 583             $hookresults = do_hook("attachment $type0/*", $links,
 584                     $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 585                     $display_filename, $where, $what);
 586         }
 587
 588         $links = $hookresults[1];
 589         $defaultlink = $hookresults[6];
 590
 591         $attachments .= '<tr><td>' .
 592             '<a href="'.$defaultlink.'">'.decodeHeader($display_filename).'</a>&nbsp;</td>' .
 593             '<td><small><b>' . show_readable_size($header->size) .
 594             '</b>&nbsp;&nbsp;</small></td>' .
 595             '<td><small>[ '.htmlspecialchars($type0).'/'.htmlspecialchars($type1).' ]&nbsp;</small></td>' .
 596             '<td><small>';
 597         $attachments .= '<b>' . $description . '</b>';
 598         $attachments .= '</small></td><td><small>&nbsp;';
 599
 600         $skipspaces = 1;
 601         foreach ($links as $val) {
 602             if ($skipspaces) {
 603                 $skipspaces = 0;
 604             } else {
 605                 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
 606             }
 607             $attachments .= '<a href="' . $val['href'] . '">'
 608                 . (isset($val['text']) && !empty($val['text']) ? $val['text'] : '')
 609                 . (isset($val['extra']) && !empty($val['extra']) ? $val['extra'] : '')
 610                 . '</a>';
 611         }
 612         unset($links);
 613         $attachments .= "</td></tr>\n";
 614     }
 615     $attachmentadd = do_hook_function('attachments_bottom',$attachments);
 616     if ($attachmentadd != '')
 617         $attachments = $attachmentadd;
 618     return $attachments;
 619 }
 620
 621 function sqimap_base64_decode(&$string) {
 622
 623     // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
 624     // fly decoding (to reduce memory usage) you have to check if the
 625     // data has incomplete pairs
 626
 627     // Remove the noise in order to check if the 4 bytes pairs are complete
 628     $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
 629
 630     $sStringRem = '';
 631     $iMod = strlen($string) % 4;
 632     if ($iMod) {
 633         $sStringRem = substr($string,-$iMod);
 634         // Check if $sStringRem contains padding characters
 635         if (substr($sStringRem,-1) != '=') {
 636             $string = substr($string,0,-$iMod);
 637         } else {
 638             $sStringRem = '';
 639         }
 640     }
 641     $string = base64_decode($string);
 642     return $sStringRem;
 643 }
 644
 645 /**
 646  * Decodes encoded message body
 647  *
 648  * This function decodes the body depending on the encoding type.
 649  * Currently quoted-printable and base64 encodings are supported.
 650  * decode_body hook was added to this function in 1.4.2/1.5.0
 651  * @param string $body encoded message body
 652  * @param string $encoding used encoding
 653  * @return string decoded string
 654  * @since 1.0
 655  */
 656 function decodeBody($body, $encoding) {
 657
 658     $body = str_replace("\r\n", "\n", $body);
 659     $encoding = strtolower($encoding);
 660
 661     $encoding_handler = do_hook_function('decode_body', $encoding);
 662
 663
 664     // plugins get first shot at decoding the body
 665     //
 666     if (!empty($encoding_handler) && function_exists($encoding_handler)) {
 667         $body = $encoding_handler('decode', $body);
 668
 669     } elseif ($encoding == 'quoted-printable' ||
 670             $encoding == 'quoted_printable') {
 671         /**
 672          * quoted_printable_decode() function is broken in older
 673          * php versions. Text with \r\n decoding was fixed only
 674          * in php 4.3.0. Minimal code requirement 4.0.4 +
 675          * str_replace("\r\n", "\n", $body); call.
 676          */
 677         $body = quoted_printable_decode($body);
 678     } elseif ($encoding == 'base64') {
 679         $body = base64_decode($body);
 680     }
 681
 682     // All other encodings are returned raw.
 683     return $body;
 684 }
 685
 686 /**
 687  * Decodes headers
 688  *
 689  * This functions decode strings that is encoded according to
 690  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 691  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 692  *
 693  * @param string $string header string that has to be made readable
 694  * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
 695  * @param boolean $htmlsave preserve spaces and sanitize html special characters. defaults to true
 696  * @param boolean $decide decide if string can be utfencoded. defaults to false
 697  * @return string decoded header string
 698  */
 699 function decodeHeader ($string, $utfencode=true,$htmlsave=true,$decide=false) {
 700     global $languages, $squirrelmail_language,$default_charset;
 701     if (is_array($string)) {
 702         $string = implode("\n", $string);
 703     }
 704
 705     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 706             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
 707         $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
 708         // Do we need to return at this point?
 709         // return $string;
 710     }
 711     $i = 0;
 712     $iLastMatch = -2;
 713     $encoded = true;
 714
 715     $aString = explode(' ',$string);
 716     $ret = '';
 717     foreach ($aString as $chunk) {
 718         if ($encoded && $chunk === '') {
 719             continue;
 720         } elseif ($chunk === '') {
 721             $ret .= ' ';
 722             continue;
 723         }
 724         $encoded = false;
 725         /* if encoded words are not separated by a linear-space-white we still catch them */
 726         $j = $i-1;
 727
 728         while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 729             /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
 730             if ($iLastMatch !== $j) {
 731                 if ($htmlsave) {
 732                     $ret .= '&#32;';
 733                 } else {
 734                     $ret .= ' ';
 735                 }
 736             }
 737             $iLastMatch = $i;
 738             $j = $i;
 739             if ($htmlsave) {
 740                 $ret .= htmlspecialchars($res[1]);
 741             } else {
 742                 $ret .= $res[1];
 743             }
 744             $encoding = ucfirst($res[3]);
 745
 746             /* decide about valid decoding */
 747             if ($decide && is_conversion_safe($res[2])) {
 748                 $utfencode=true;
 749                 $can_be_encoded=true;
 750             } else {
 751                 $can_be_encoded=false;
 752             }
 753             switch ($encoding)
 754             {
 755                 case 'B':
 756                     $replace = base64_decode($res[4]);
 757                     if ($utfencode) {
 758                         if ($can_be_encoded) {
 759                             /* convert string to different charset,
 760                              * if functions asks for it (usually in compose)
 761                              */
 762                             $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsave);
 763                         } else {
 764                             // convert string to html codes in order to display it
 765                             $ret .= charset_decode($res[2],$replace);
 766                         }
 767                     } else {
 768                         if ($htmlsave) {
 769                             $replace = htmlspecialchars($replace);
 770                         }
 771                         $ret.= $replace;
 772                     }
 773                     break;
 774                 case 'Q':
 775                     $replace = str_replace('_', ' ', $res[4]);
 776                     $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 777                             $replace);
 778                     if ($utfencode) {
 779                         if ($can_be_encoded) {
 780                             /* convert string to different charset,
 781                              * if functions asks for it (usually in compose)
 782                              */
 783                             $replace = charset_convert($res[2], $replace,$default_charset,$htmlsave);
 784                         } else {
 785                             // convert string to html codes in order to display it
 786                             $replace = charset_decode($res[2], $replace);
 787                         }
 788                     } else {
 789                         if ($htmlsave) {
 790                             $replace = htmlspecialchars($replace);
 791                         }
 792                     }
 793                     $ret .= $replace;
 794                     break;
 795                 default:
 796                     break;
 797             }
 798             $chunk = $res[5];
 799             $encoded = true;
 800         }
 801         if (!$encoded) {
 802             if ($htmlsave) {
 803                 $ret .= '&#32;';
 804             } else {
 805                 $ret .= ' ';
 806             }
 807         }
 808
 809         if (!$encoded && $htmlsave) {
 810             $ret .= htmlspecialchars($chunk);
 811         } else {
 812             $ret .= $chunk;
 813         }
 814         ++$i;
 815     }
 816     /* remove the first added space */
 817     if ($ret) {
 818         if ($htmlsave) {
 819             $ret = substr($ret,5);
 820         } else {
 821             $ret = substr($ret,1);
 822         }
 823     }
 824
 825     return $ret;
 826 }
 827
 828 /**
 829  * Encodes header
 830  *
 831  * Function uses XTRA_CODE _encodeheader function, if such function exists.
 832  *
 833  * Function uses Q encoding by default and encodes a string according to RFC
 834  * 1522 for use in headers if it contains 8-bit characters or anything that
 835  * looks like it should be encoded.
 836  *
 837  * Function switches to B encoding and encodeHeaderBase64() function, if
 838  * string is 8bit and multibyte character set supported by mbstring extension
 839  * is used. It can cause E_USER_NOTICE errors, if interface is used with
 840  * multibyte character set unsupported by mbstring extension.
 841  *
 842  * @param string $string header string, that has to be encoded
 843  * @return string quoted-printable encoded string
 844  * @todo make $mb_charsets system wide constant
 845  */
 846 function encodeHeader ($string) {
 847     global $default_charset, $languages, $squirrelmail_language;
 848
 849     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 850             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
 851         return  call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
 852     }
 853
 854     // Use B encoding for multibyte charsets
 855     $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
 856     if (in_array($default_charset,$mb_charsets) &&
 857         in_array($default_charset,sq_mb_list_encodings()) &&
 858         sq_is8bit($string)) {
 859         return encodeHeaderBase64($string,$default_charset);
 860     } elseif (in_array($default_charset,$mb_charsets) &&
 861               sq_is8bit($string) &&
 862               ! in_array($default_charset,sq_mb_list_encodings())) {
 863         // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
 864         // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
 865     }
 866
 867     // Encode only if the string contains 8-bit characters or =?
 868     $j = strlen($string);
 869     $max_l = 75 - strlen($default_charset) - 7;
 870     $aRet = array();
 871     $ret = '';
 872     $iEncStart = $enc_init = false;
 873     $cur_l = $iOffset = 0;
 874     for($i = 0; $i < $j; ++$i) {
 875         switch($string{$i})
 876         {
 877             case '=':
 878             case '<':
 879             case '>':
 880             case ',':
 881             case '?':
 882             case '_':
 883                 if ($iEncStart === false) {
 884                     $iEncStart = $i;
 885                 }
 886                 $cur_l+=3;
 887                 if ($cur_l > ($max_l-2)) {
 888                     /* if there is an stringpart that doesn't need encoding, add it */
 889                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 890                     $aRet[] = "=?$default_charset?Q?$ret?=";
 891                     $iOffset = $i;
 892                     $cur_l = 0;
 893                     $ret = '';
 894                     $iEncStart = false;
 895                 } else {
 896                     $ret .= sprintf("=%02X",ord($string{$i}));
 897                 }
 898                 break;
 899             case '(':
 900             case ')':
 901                 if ($iEncStart !== false) {
 902                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 903                     $aRet[] = "=?$default_charset?Q?$ret?=";
 904                     $iOffset = $i;
 905                     $cur_l = 0;
 906                     $ret = '';
 907                     $iEncStart = false;
 908                 }
 909                 break;
 910             case ' ':
 911                 if ($iEncStart !== false) {
 912                     $cur_l++;
 913                     if ($cur_l > $max_l) {
 914                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 915                         $aRet[] = "=?$default_charset?Q?$ret?=";
 916                         $iOffset = $i;
 917                         $cur_l = 0;
 918                         $ret = '';
 919                         $iEncStart = false;
 920                     } else {
 921                         $ret .= '_';
 922                     }
 923                 }
 924                 break;
 925             default:
 926                 $k = ord($string{$i});
 927                 if ($k > 126) {
 928                     if ($iEncStart === false) {
 929                         // do not start encoding in the middle of a string, also take the rest of the word.
 930                         $sLeadString = substr($string,0,$i);
 931                         $aLeadString = explode(' ',$sLeadString);
 932                         $sToBeEncoded = array_pop($aLeadString);
 933                         $iEncStart = $i - strlen($sToBeEncoded);
 934                         $ret .= $sToBeEncoded;
 935                         $cur_l += strlen($sToBeEncoded);
 936                     }
 937                     $cur_l += 3;
 938                     /* first we add the encoded string that reached it's max size */
 939                     if ($cur_l > ($max_l-2)) {
 940                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 941                         $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
 942                         $cur_l = 3;
 943                         $ret = '';
 944                         $iOffset = $i;
 945                         $iEncStart = $i;
 946                     }
 947                     $enc_init = true;
 948                     $ret .= sprintf("=%02X", $k);
 949                 } else {
 950                     if ($iEncStart !== false) {
 951                         $cur_l++;
 952                         if ($cur_l > $max_l) {
 953                             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 954                             $aRet[] = "=?$default_charset?Q?$ret?=";
 955                             $iEncStart = false;
 956                             $iOffset = $i;
 957                             $cur_l = 0;
 958                             $ret = '';
 959                         } else {
 960                             $ret .= $string{$i};
 961                         }
 962                     }
 963                 }
 964                 break;
 965         }
 966     }
 967
 968     if ($enc_init) {
 969         if ($iEncStart !== false) {
 970             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 971             $aRet[] = "=?$default_charset?Q?$ret?=";
 972         } else {
 973             $aRet[] = substr($string,$iOffset);
 974         }
 975         $string = implode('',$aRet);
 976     }
 977     return $string;
 978 }
 979
 980 /**
 981  * Encodes string according to rfc2047 B encoding header formating rules
 982  *
 983  * It is recommended way to encode headers with character sets that store
 984  * symbols in more than one byte.
 985  *
 986  * Function requires mbstring support. If required mbstring functions are missing,
 987  * function returns false and sets E_USER_WARNING level error message.
 988  *
 989  * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
 990  * that mbstring functions will generate E_WARNING errors, if unsupported
 991  * character set is used. mb_encode_mimeheader function provided by php
 992  * mbstring extension is not used in order to get better control of header
 993  * encoding.
 994  *
 995  * Used php code functions - function_exists(), trigger_error(), strlen()
 996  * (is used with charset names and base64 strings). Used php mbstring
 997  * functions - mb_strlen and mb_substr.
 998  *
 999  * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
1000  * encoding), rfc 2822 (header folding)
1001  *
1002  * @param string $string header string that must be encoded
1003  * @param string $charset character set. Must be supported by mbstring extension.
1004  * Use sq_mb_list_encodings() to detect supported charsets.
1005  * @return string string encoded according to rfc2047 B encoding formating rules
1006  * @since 1.5.1
1007  * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1008  * @todo Do we want to control max length of header?
1009  * @todo Do we want to control EOL (end-of-line) marker?
1010  * @todo Do we want to translate error message?
1011  */
1012 function encodeHeaderBase64($string,$charset) {
1013     /**
1014      * Check mbstring function requirements.
1015      */
1016     if (! function_exists('mb_strlen') ||
1017         ! function_exists('mb_substr')) {
1018         // set E_USER_WARNING
1019         trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
1020         // return false
1021         return false;
1022     }
1023
1024     // initial return array
1025     $aRet = array();
1026
1027     /**
1028      * header length = 75 symbols max (same as in encodeHeader)
1029      * remove $charset length
1030      * remove =? ? ?= (5 chars)
1031      * remove 2 more chars (\r\n ?)
1032      */
1033     $iMaxLength = 75 - strlen($charset) - 7;
1034
1035     // set first character position
1036     $iStartCharNum = 0;
1037
1038     // loop through all characters. count characters and not bytes.
1039     for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1040         // encode string from starting character to current character.
1041         $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1042
1043         // Check encoded string length
1044         if(strlen($encoded_string)>$iMaxLength) {
1045             // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1046             $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1047
1048             // set new starting character
1049             $iStartCharNum = $iCharNum-1;
1050
1051             // encode last char (in case it is last character in string)
1052             $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1053         } // if string is shorter than max length - add next character
1054     }
1055
1056     // add last encoded string to array
1057     $aRet[] = $encoded_string;
1058
1059     // set initial return string
1060     $sRet = '';
1061
1062     // loop through encoded strings
1063     foreach($aRet as $string) {
1064         // TODO: Do we want to control EOL (end-of-line) marker
1065         if ($sRet!='') $sRet.= " ";
1066
1067         // add header tags and encoded string to return string
1068         $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1069     }
1070
1071     return $sRet;
1072 }
1073
1074 /* This function trys to locate the entity_id of a specific mime element */
1075 function find_ent_id($id, $message) {
1076     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1077         if ($message->entities[$i]->header->type0 == 'multipart')  {
1078             $ret = find_ent_id($id, $message->entities[$i]);
1079         } else {
1080             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
1081 //                if (sq_check_save_extension($message->entities[$i])) {
1082                 return $message->entities[$i]->entity_id;
1083 //                }
1084             } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1085                 /**
1086                  * This is part of a fix for Outlook Express 6.x generating
1087                  * cid URLs without creating content-id headers
1088                  * @@JA - 20050207
1089                  */
1090                 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1091                     return $message->entities[$i]->entity_id;
1092                 }
1093             }
1094         }
1095     }
1096     return $ret;
1097 }
1098
1099 function sq_check_save_extension($message) {
1100     $filename = $message->getFilename();
1101     $ext = substr($filename, strrpos($filename,'.')+1);
1102     $save_extensions = array('jpg','jpeg','gif','png','bmp');
1103     return in_array($ext, $save_extensions);
1104 }
1105
1106
1107 /**
1108  ** HTMLFILTER ROUTINES
1109  */
1110
1111 /**
1112  * This function checks attribute values for entity-encoded values
1113  * and returns them translated into 8-bit strings so we can run
1114  * checks on them.
1115  *
1116  * @param  $attvalue A string to run entity check against.
1117  * @return           Nothing, modifies a reference value.
1118  */
1119 function sq_defang(&$attvalue){
1120     $me = 'sq_defang';
1121     /**
1122      * Skip this if there aren't ampersands or backslashes.
1123      */
1124     if (strpos($attvalue, '&') === false
1125         && strpos($attvalue, '\\') === false){
1126         return;
1127     }
1128     $m = false;
1129     do {
1130         $m = false;
1131         $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1132         $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1133         $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1134     } while ($m == true);
1135     $attvalue = stripslashes($attvalue);
1136 }
1137
1138 /**
1139  * Kill any tabs, newlines, or carriage returns. Our friends the
1140  * makers of the browser with 95% market value decided that it'd
1141  * be funny to make "java[tab]script" be just as good as "javascript".
1142  *
1143  * @param  attvalue  The attribute value before extraneous spaces removed.
1144  * @return attvalue  Nothing, modifies a reference value.
1145  */
1146 function sq_unspace(&$attvalue){
1147     $me = 'sq_unspace';
1148     if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1149         $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1150                                 Array('',   '',   '',   '',   ''), $attvalue);
1151     }
1152 }
1153
1154 /**
1155  * This function returns the final tag out of the tag name, an array
1156  * of attributes, and the type of the tag. This function is called by
1157  * sq_sanitize internally.
1158  *
1159  * @param  $tagname  the name of the tag.
1160  * @param  $attary   the array of attributes and their values
1161  * @param  $tagtype  The type of the tag (see in comments).
1162  * @return           a string with the final tag representation.
1163  */
1164 function sq_tagprint($tagname, $attary, $tagtype){
1165     $me = 'sq_tagprint';
1166
1167     if ($tagtype == 2){
1168         $fulltag = '</' . $tagname . '>';
1169     } else {
1170         $fulltag = '<' . $tagname;
1171         if (is_array($attary) && sizeof($attary)){
1172             $atts = Array();
1173             while (list($attname, $attvalue) = each($attary)){
1174                 array_push($atts, "$attname=$attvalue");
1175             }
1176             $fulltag .= ' ' . join(" ", $atts);
1177         }
1178         if ($tagtype == 3){
1179             $fulltag .= ' /';
1180         }
1181         $fulltag .= '>';
1182     }
1183     return $fulltag;
1184 }
1185
1186 /**
1187  * A small helper function to use with array_walk. Modifies a by-ref
1188  * value and makes it lowercase.
1189  *
1190  * @param  $val a value passed by-ref.
1191  * @return      void since it modifies a by-ref value.
1192  */
1193 function sq_casenormalize(&$val){
1194     $val = strtolower($val);
1195 }
1196
1197 /**
1198  * This function skips any whitespace from the current position within
1199  * a string and to the next non-whitespace value.
1200  *
1201  * @param  $body   the string
1202  * @param  $offset the offset within the string where we should start
1203  *                 looking for the next non-whitespace character.
1204  * @return         the location within the $body where the next
1205  *                 non-whitespace char is located.
1206  */
1207 function sq_skipspace($body, $offset){
1208     $me = 'sq_skipspace';
1209     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1210     if (sizeof($matches{1})){
1211         $count = strlen($matches{1});
1212         $offset += $count;
1213     }
1214     return $offset;
1215 }
1216
1217 /**
1218  * This function looks for the next character within a string.  It's
1219  * really just a glorified "strpos", except it catches if failures
1220  * nicely.
1221  *
1222  * @param  $body   The string to look for needle in.
1223  * @param  $offset Start looking from this position.
1224  * @param  $needle The character/string to look for.
1225  * @return         location of the next occurance of the needle, or
1226  *                 strlen($body) if needle wasn't found.
1227  */
1228 function sq_findnxstr($body, $offset, $needle){
1229     $me  = 'sq_findnxstr';
1230     $pos = strpos($body, $needle, $offset);
1231     if ($pos === FALSE){
1232         $pos = strlen($body);
1233     }
1234     return $pos;
1235 }
1236
1237 /**
1238  * This function takes a PCRE-style regexp and tries to match it
1239  * within the string.
1240  *
1241  * @param  $body   The string to look for needle in.
1242  * @param  $offset Start looking from here.
1243  * @param  $reg    A PCRE-style regex to match.
1244  * @return         Returns a false if no matches found, or an array
1245  *                 with the following members:
1246  *                 - integer with the location of the match within $body
1247  *                 - string with whatever content between offset and the match
1248  *                 - string with whatever it is we matched
1249  */
1250 function sq_findnxreg($body, $offset, $reg){
1251     $me = 'sq_findnxreg';
1252     $matches = Array();
1253     $retarr = Array();
1254     preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1255     if (!isset($matches{0}) || !$matches{0}){
1256         $retarr = false;
1257     } else {
1258         $retarr{0} = $offset + strlen($matches{1});
1259         $retarr{1} = $matches{1};
1260         $retarr{2} = $matches{2};
1261     }
1262     return $retarr;
1263 }
1264
1265 /**
1266  * This function looks for the next tag.
1267  *
1268  * @param  $body   String where to look for the next tag.
1269  * @param  $offset Start looking from here.
1270  * @return         false if no more tags exist in the body, or
1271  *                 an array with the following members:
1272  *                 - string with the name of the tag
1273  *                 - array with attributes and their values
1274  *                 - integer with tag type (1, 2, or 3)
1275  *                 - integer where the tag starts (starting "<")
1276  *                 - integer where the tag ends (ending ">")
1277  *                 first three members will be false, if the tag is invalid.
1278  */
1279 function sq_getnxtag($body, $offset){
1280     $me = 'sq_getnxtag';
1281     if ($offset > strlen($body)){
1282         return false;
1283     }
1284     $lt = sq_findnxstr($body, $offset, "<");
1285     if ($lt == strlen($body)){
1286         return false;
1287     }
1288     /**
1289      * We are here:
1290      * blah blah <tag attribute="value">
1291      * \---------^
1292      */
1293     $pos = sq_skipspace($body, $lt+1);
1294     if ($pos >= strlen($body)){
1295         return Array(false, false, false, $lt, strlen($body));
1296     }
1297     /**
1298      * There are 3 kinds of tags:
1299      * 1. Opening tag, e.g.:
1300      *    <a href="blah">
1301      * 2. Closing tag, e.g.:
1302      *    </a>
1303      * 3. XHTML-style content-less tag, e.g.:
1304      *    <img src="blah" />
1305      */
1306     $tagtype = false;
1307     switch (substr($body, $pos, 1)){
1308         case '/':
1309             $tagtype = 2;
1310             $pos++;
1311             break;
1312         case '!':
1313             /**
1314              * A comment or an SGML declaration.
1315              */
1316             if (substr($body, $pos+1, 2) == "--"){
1317                 $gt = strpos($body, "-->", $pos);
1318                 if ($gt === false){
1319                     $gt = strlen($body);
1320                 } else {
1321                     $gt += 2;
1322                 }
1323                 return Array(false, false, false, $lt, $gt);
1324             } else {
1325                 $gt = sq_findnxstr($body, $pos, ">");
1326                 return Array(false, false, false, $lt, $gt);
1327             }
1328             break;
1329         default:
1330             /**
1331              * Assume tagtype 1 for now. If it's type 3, we'll switch values
1332              * later.
1333              */
1334             $tagtype = 1;
1335             break;
1336     }
1337
1338     $tag_start = $pos;
1339     $tagname = '';
1340     /**
1341      * Look for next [\W-_], which will indicate the end of the tag name.
1342      */
1343     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1344     if ($regary == false){
1345         return Array(false, false, false, $lt, strlen($body));
1346     }
1347     list($pos, $tagname, $match) = $regary;
1348     $tagname = strtolower($tagname);
1349
1350     /**
1351      * $match can be either of these:
1352      * '>'  indicating the end of the tag entirely.
1353      * '\s' indicating the end of the tag name.
1354      * '/'  indicating that this is type-3 xhtml tag.
1355      *
1356      * Whatever else we find there indicates an invalid tag.
1357      */
1358     switch ($match){
1359         case '/':
1360             /**
1361              * This is an xhtml-style tag with a closing / at the
1362              * end, like so: <img src="blah" />. Check if it's followed
1363              * by the closing bracket. If not, then this tag is invalid
1364              */
1365             if (substr($body, $pos, 2) == "/>"){
1366                 $pos++;
1367                 $tagtype = 3;
1368             } else {
1369                 $gt = sq_findnxstr($body, $pos, ">");
1370                 $retary = Array(false, false, false, $lt, $gt);
1371                 return $retary;
1372             }
1373         case '>':
1374             return Array($tagname, false, $tagtype, $lt, $pos);
1375             break;
1376         default:
1377             /**
1378              * Check if it's whitespace
1379              */
1380             if (!preg_match('/\s/', $match)){
1381                 /**
1382                  * This is an invalid tag! Look for the next closing ">".
1383                  */
1384                 $gt = sq_findnxstr($body, $lt, ">");
1385                 return Array(false, false, false, $lt, $gt);
1386             }
1387             break;
1388     }
1389
1390     /**
1391      * At this point we're here:
1392      * <tagname  attribute='blah'>
1393      * \-------^
1394      *
1395      * At this point we loop in order to find all attributes.
1396      */
1397     $attname = '';
1398     $atttype = false;
1399     $attary = Array();
1400
1401     while ($pos <= strlen($body)){
1402         $pos = sq_skipspace($body, $pos);
1403         if ($pos == strlen($body)){
1404             /**
1405              * Non-closed tag.
1406              */
1407             return Array(false, false, false, $lt, $pos);
1408         }
1409         /**
1410          * See if we arrived at a ">" or "/>", which means that we reached
1411          * the end of the tag.
1412          */
1413         $matches = Array();
1414         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1415             /**
1416              * Yep. So we did.
1417              */
1418             $pos += strlen($matches{1});
1419             if ($matches{2} == "/>"){
1420                 $tagtype = 3;
1421                 $pos++;
1422             }
1423             return Array($tagname, $attary, $tagtype, $lt, $pos);
1424         }
1425
1426         /**
1427          * There are several types of attributes, with optional
1428          * [:space:] between members.
1429          * Type 1:
1430          *   attrname[:space:]=[:space:]'CDATA'
1431          * Type 2:
1432          *   attrname[:space:]=[:space:]"CDATA"
1433          * Type 3:
1434          *   attr[:space:]=[:space:]CDATA
1435          * Type 4:
1436          *   attrname
1437          *
1438          * We leave types 1 and 2 the same, type 3 we check for
1439          * '"' and convert to "&quot" if needed, then wrap in
1440          * double quotes. Type 4 we convert into:
1441          * attrname="yes".
1442          */
1443         $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1444         if ($regary == false){
1445             /**
1446              * Looks like body ended before the end of tag.
1447              */
1448             return Array(false, false, false, $lt, strlen($body));
1449         }
1450         list($pos, $attname, $match) = $regary;
1451         $attname = strtolower($attname);
1452         /**
1453          * We arrived at the end of attribute name. Several things possible
1454          * here:
1455          * '>'  means the end of the tag and this is attribute type 4
1456          * '/'  if followed by '>' means the same thing as above
1457          * '\s' means a lot of things -- look what it's followed by.
1458          *      anything else means the attribute is invalid.
1459          */
1460         switch($match){
1461             case '/':
1462                 /**
1463                  * This is an xhtml-style tag with a closing / at the
1464                  * end, like so: <img src="blah" />. Check if it's followed
1465                  * by the closing bracket. If not, then this tag is invalid
1466                  */
1467                 if (substr($body, $pos, 2) == "/>"){
1468                     $pos++;
1469                     $tagtype = 3;
1470                 } else {
1471                     $gt = sq_findnxstr($body, $pos, ">");
1472                     $retary = Array(false, false, false, $lt, $gt);
1473                     return $retary;
1474                 }
1475             case '>':
1476                 $attary{$attname} = '"yes"';
1477                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1478                 break;
1479             default:
1480                 /**
1481                  * Skip whitespace and see what we arrive at.
1482                  */
1483                 $pos = sq_skipspace($body, $pos);
1484                 $char = substr($body, $pos, 1);
1485                 /**
1486                  * Two things are valid here:
1487                  * '=' means this is attribute type 1 2 or 3.
1488                  * \w means this was attribute type 4.
1489                  * anything else we ignore and re-loop. End of tag and
1490                  * invalid stuff will be caught by our checks at the beginning
1491                  * of the loop.
1492                  */
1493                 if ($char == "="){
1494                     $pos++;
1495                     $pos = sq_skipspace($body, $pos);
1496                     /**
1497                      * Here are 3 possibilities:
1498                      * "'"  attribute type 1
1499                      * '"'  attribute type 2
1500                      * everything else is the content of tag type 3
1501                      */
1502                     $quot = substr($body, $pos, 1);
1503                     if ($quot == "'"){
1504                         $regary = sq_findnxreg($body, $pos+1, "\'");
1505                         if ($regary == false){
1506                             return Array(false, false, false, $lt, strlen($body));
1507                         }
1508                         list($pos, $attval, $match) = $regary;
1509                         $pos++;
1510                         $attary{$attname} = "'" . $attval . "'";
1511                     } else if ($quot == '"'){
1512                         $regary = sq_findnxreg($body, $pos+1, '\"');
1513                         if ($regary == false){
1514                             return Array(false, false, false, $lt, strlen($body));
1515                         }
1516                         list($pos, $attval, $match) = $regary;
1517                         $pos++;
1518                         $attary{$attname} = '"' . $attval . '"';
1519                     } else {
1520                         /**
1521                          * These are hateful. Look for \s, or >.
1522                          */
1523                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1524                         if ($regary == false){
1525                             return Array(false, false, false, $lt, strlen($body));
1526                         }
1527                         list($pos, $attval, $match) = $regary;
1528                         /**
1529                          * If it's ">" it will be caught at the top.
1530                          */
1531                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1532                         $attary{$attname} = '"' . $attval . '"';
1533                     }
1534                 } else if (preg_match("|[\w/>]|", $char)) {
1535                     /**
1536                      * That was attribute type 4.
1537                      */
1538                     $attary{$attname} = '"yes"';
1539                 } else {
1540                     /**
1541                      * An illegal character. Find next '>' and return.
1542                      */
1543                     $gt = sq_findnxstr($body, $pos, ">");
1544                     return Array(false, false, false, $lt, $gt);
1545                 }
1546                 break;
1547         }
1548     }
1549     /**
1550      * The fact that we got here indicates that the tag end was never
1551      * found. Return invalid tag indication so it gets stripped.
1552      */
1553     return Array(false, false, false, $lt, strlen($body));
1554 }
1555
1556 /**
1557  * Translates entities into literal values so they can be checked.
1558  *
1559  * @param $attvalue the by-ref value to check.
1560  * @param $regex    the regular expression to check against.
1561  * @param $hex      whether the entites are hexadecimal.
1562  * @return          True or False depending on whether there were matches.
1563  */
1564 function sq_deent(&$attvalue, $regex, $hex=false){
1565     $me = 'sq_deent';
1566     $ret_match = false;
1567     preg_match_all($regex, $attvalue, $matches);
1568     if (is_array($matches) && sizeof($matches[0]) > 0){
1569         $repl = Array();
1570         for ($i = 0; $i < sizeof($matches[0]); $i++){
1571             $numval = $matches[1][$i];
1572             if ($hex){
1573                 $numval = hexdec($numval);
1574             }
1575             $repl{$matches[0][$i]} = chr($numval);
1576         }
1577         $attvalue = strtr($attvalue, $repl);
1578         return true;
1579     } else {
1580         return false;
1581     }
1582 }
1583
1584 /**
1585  * This function runs various checks against the attributes.
1586  *
1587  * @param  $tagname         String with the name of the tag.
1588  * @param  $attary          Array with all tag attributes.
1589  * @param  $rm_attnames     See description for sq_sanitize
1590  * @param  $bad_attvals     See description for sq_sanitize
1591  * @param  $add_attr_to_tag See description for sq_sanitize
1592  * @param  $message         message object
1593  * @param  $id              message id
1594  * @return                  Array with modified attributes.
1595  */
1596 function sq_fixatts($tagname,
1597                     $attary,
1598                     $rm_attnames,
1599                     $bad_attvals,
1600                     $add_attr_to_tag,
1601                     $message,
1602                     $id,
1603                     $mailbox
1604                     ){
1605     $me = 'sq_fixatts';
1606     while (list($attname, $attvalue) = each($attary)){
1607         /**
1608          * See if this attribute should be removed.
1609          */
1610         foreach ($rm_attnames as $matchtag=>$matchattrs){
1611             if (preg_match($matchtag, $tagname)){
1612                 foreach ($matchattrs as $matchattr){
1613                     if (preg_match($matchattr, $attname)){
1614                         unset($attary{$attname});
1615                         continue;
1616                     }
1617                 }
1618             }
1619         }
1620         /**
1621          * Remove any backslashes, entities, and extraneous whitespace.
1622          */
1623         sq_defang($attvalue);
1624         sq_unspace($attvalue);
1625
1626         /**
1627          * Now let's run checks on the attvalues.
1628          * I don't expect anyone to comprehend this. If you do,
1629          * get in touch with me so I can drive to where you live and
1630          * shake your hand personally. :)
1631          */
1632         foreach ($bad_attvals as $matchtag=>$matchattrs){
1633             if (preg_match($matchtag, $tagname)){
1634                 foreach ($matchattrs as $matchattr=>$valary){
1635                     if (preg_match($matchattr, $attname)){
1636                         /**
1637                          * There are two arrays in valary.
1638                          * First is matches.
1639                          * Second one is replacements
1640                          */
1641                         list($valmatch, $valrepl) = $valary;
1642                         $newvalue =
1643                             preg_replace($valmatch, $valrepl, $attvalue);
1644                         if ($newvalue != $attvalue){
1645                             $attary{$attname} = $newvalue;
1646                         }
1647                     }
1648                 }
1649             }
1650         }
1651
1652         /**
1653          * Replace empty src tags with the blank image.  src is only used
1654          * for frames, images, and image inputs.  Doing a replace should
1655          * not affect them working as should be, however it will stop
1656          * IE from being kicked off when src for img tags are not set
1657          */
1658         if (($attname == 'src') && ($attvalue == '""')) {
1659             $attary{$attname} = '"' . SM_PATH . 'images/blank.png"';
1660         }
1661
1662         /**
1663          * Turn cid: urls into http-friendly ones.
1664          */
1665         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1666             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1667         }
1668
1669         /**
1670          * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1671          * One day MS might actually make it match something useful, for now, falling
1672          * back to using cid2http, so we can grab the blank.png.
1673          */
1674         if (preg_match("/^[\'\"]\s*outbind:\/\//si", $attvalue)) {
1675             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1676         }
1677
1678     }
1679     /**
1680      * See if we need to append any attributes to this tag.
1681      */
1682     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1683         if (preg_match($matchtag, $tagname)){
1684             $attary = array_merge($attary, $addattary);
1685         }
1686     }
1687     return $attary;
1688 }
1689
1690 /**
1691  * This function edits the style definition to make them friendly and
1692  * usable in SquirrelMail.
1693  *
1694  * @param  $message  the message object
1695  * @param  $id       the message id
1696  * @param  $content  a string with whatever is between <style> and </style>
1697  * @param  $mailbox  the message mailbox
1698  * @return           a string with edited content.
1699  */
1700 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1701     global $view_unsafe_images;
1702     $me = 'sq_fixstyle';
1703     $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1704     if ($ret == FALSE){
1705         return array(FALSE, strlen($body));
1706     }
1707     $newpos = $ret[0] + strlen($ret[2]);
1708     $content = $ret[1];
1709     /**
1710     * First look for general BODY style declaration, which would be
1711     * like so:
1712     * body {background: blah-blah}
1713     * and change it to .bodyclass so we can just assign it to a <div>
1714     */
1715     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1716     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1717     /**
1718     * Fix url('blah') declarations.
1719     */
1720     //   $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1721     //                           "url(\\1$secremoveimg\\2)", $content);
1722     // remove NUL
1723     $content = str_replace("\0", "", $content);
1724     // translate ur\l and variations (IE parses that)
1725     $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
1726     // NB I insert NUL characters to keep to avoid an infinite loop. They are removed after the loop.
1727     while (preg_match("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si", $content, $matches)) {
1728         $sProto = strtolower($matches[1]);
1729         switch ($sProto) {
1730             /**
1731              * Fix url('https*://.*) declarations but only if $view_unsafe_images
1732              * is false.
1733              */
1734             case 'https':
1735             case 'http':
1736                 if (!$view_unsafe_images){
1737
1738                     $sExpr = "/url\s*\(\s*[\'\"]?\s*$sProto*:.*[\'\"]?\s*\)/si";
1739                     $content = preg_replace($sExpr, "u\0r\0l(\\1$secremoveimg\\2)", $content);
1740
1741                 } else {
1742                     $content = preg_replace('/url/i',"u\0r\0l",$content);
1743                 }
1744                 break;
1745             /**
1746              * Fix urls that refer to cid:
1747              */
1748             case 'cid':
1749                 $cidurl = 'cid:'. $matches[2];
1750                 $httpurl = sq_cid2http($message, $id, $cidurl, $mailbox);
1751                 // escape parentheses that can modify the regular expression
1752                 $cidurl = str_replace(array('(',')'),array('\\(','\\)'),$cidurl);
1753                 $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1754                                         "u\0r\0l($httpurl)", $content);
1755                 break;
1756             default:
1757                 /**
1758                  * replace url with protocol other then the white list
1759                  * http,https and cid by an empty string.
1760                  */
1761                 $content = preg_replace("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si",
1762                                  "", $content);
1763                 break;
1764         }
1765     }
1766     // remove NUL
1767     $content = str_replace("\0", "", $content);
1768    /**
1769     * Remove any backslashes, entities, and extraneous whitespace.
1770     */
1771     $contentTemp = $content;
1772     sq_defang($contentTemp);
1773     sq_unspace($contentTemp);
1774
1775     /**
1776      * Fix stupid css declarations which lead to vulnerabilities
1777      * in IE.
1778      */
1779     $match   = Array('/\/\*.*\*\//',
1780                     '/expression/i',
1781                     '/behaviou*r/i',
1782                     '/binding/i',
1783                     '/include-source/i');
1784     $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy');
1785     $contentNew = preg_replace($match, $replace, $contentTemp);
1786     if ($contentNew !== $contentTemp) {
1787         // insecure css declarations are used. From now on we don't care
1788         // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
1789         $content = $contentNew;
1790     }
1791     return array($content, $newpos);
1792 }
1793
1794
1795 /**
1796  * This function converts cid: url's into the ones that can be viewed in
1797  * the browser.
1798  *
1799  * @param  $message  the message object
1800  * @param  $id       the message id
1801  * @param  $cidurl   the cid: url.
1802  * @param  $mailbox  the message mailbox
1803  * @return           a string with a http-friendly url
1804  */
1805 function sq_cid2http($message, $id, $cidurl, $mailbox){
1806     /**
1807      * Get rid of quotes.
1808      */
1809     $quotchar = substr($cidurl, 0, 1);
1810     if ($quotchar == '"' || $quotchar == "'"){
1811         $cidurl = str_replace($quotchar, "", $cidurl);
1812     } else {
1813         $quotchar = '';
1814     }
1815     $cidurl = substr(trim($cidurl), 4);
1816
1817     $match_str = '/\{.*?\}\//';
1818     $str_rep = '';
1819     $cidurl = preg_replace($match_str, $str_rep, $cidurl);
1820
1821     $linkurl = find_ent_id($cidurl, $message);
1822     /* in case of non-save cid links $httpurl should be replaced by a sort of
1823        unsave link image */
1824     $httpurl = '';
1825
1826     /**
1827      * This is part of a fix for Outlook Express 6.x generating
1828      * cid URLs without creating content-id headers. These images are
1829      * not part of the multipart/related html mail. The html contains
1830      * <img src="cid:{some_id}/image_filename.ext"> references to
1831      * attached images with as goal to render them inline although
1832      * the attachment disposition property is not inline.
1833      */
1834
1835     if (empty($linkurl)) {
1836         if (preg_match('/{.*}\//', $cidurl)) {
1837             $cidurl = preg_replace('/{.*}\//','', $cidurl);
1838             if (!empty($cidurl)) {
1839                 $linkurl = find_ent_id($cidurl, $message);
1840             }
1841         }
1842     }
1843
1844     if (!empty($linkurl)) {
1845         $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1846             "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1847             '&amp;ent_id=' . $linkurl . $quotchar;
1848     } else {
1849         /**
1850          * If we couldn't generate a proper img url, drop in a blank image
1851          * instead of sending back empty, otherwise it causes unusual behaviour
1852          */
1853         $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
1854     }
1855
1856     return $httpurl;
1857 }
1858
1859 /**
1860  * This function changes the <body> tag into a <div> tag since we
1861  * can't really have a body-within-body.
1862  *
1863  * @param  $attary   an array of attributes and values of <body>
1864  * @param  $mailbox  mailbox we're currently reading (for cid2http)
1865  * @param  $message  current message (for cid2http)
1866  * @param  $id       current message id (for cid2http)
1867  * @return           a modified array of attributes to be set for <div>
1868  */
1869 function sq_body2div($attary, $mailbox, $message, $id){
1870     $me = 'sq_body2div';
1871     $divattary = Array('class' => "'bodyclass'");
1872     $text = '#000000';
1873     $has_bgc_stl = $has_txt_stl = false;
1874     $styledef = '';
1875     if (is_array($attary) && sizeof($attary) > 0){
1876         foreach ($attary as $attname=>$attvalue){
1877             $quotchar = substr($attvalue, 0, 1);
1878             $attvalue = str_replace($quotchar, "", $attvalue);
1879             switch ($attname){
1880                 case 'background':
1881                     $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1882                     $styledef .= "background-image: url('$attvalue'); ";
1883                     break;
1884                 case 'bgcolor':
1885                     $has_bgc_stl = true;
1886                     $styledef .= "background-color: $attvalue; ";
1887                     break;
1888                 case 'text':
1889                     $has_txt_stl = true;
1890                     $styledef .= "color: $attvalue; ";
1891                     break;
1892             }
1893         }
1894         // Outlook defines a white bgcolor and no text color. This can lead to
1895         // white text on a white bg with certain themes.
1896         if ($has_bgc_stl && !$has_txt_stl) {
1897             $styledef .= "color: $text; ";
1898         }
1899         if (strlen($styledef) > 0){
1900             $divattary{"style"} = "\"$styledef\"";
1901         }
1902     }
1903     return $divattary;
1904 }
1905
1906 /**
1907  * This is the main function and the one you should actually be calling.
1908  * There are several variables you should be aware of an which need
1909  * special description.
1910  *
1911  * Since the description is quite lengthy, see it here:
1912  * http://linux.duke.edu/projects/mini/htmlfilter/
1913  *
1914  * @param $body                 the string with HTML you wish to filter
1915  * @param $tag_list             see description above
1916  * @param $rm_tags_with_content see description above
1917  * @param $self_closing_tags    see description above
1918  * @param $force_tag_closing    see description above
1919  * @param $rm_attnames          see description above
1920  * @param $bad_attvals          see description above
1921  * @param $add_attr_to_tag      see description above
1922  * @param $message              message object
1923  * @param $id                   message id
1924  * @return                      sanitized html safe to show on your pages.
1925  */
1926 function sq_sanitize($body,
1927                      $tag_list,
1928                      $rm_tags_with_content,
1929                      $self_closing_tags,
1930                      $force_tag_closing,
1931                      $rm_attnames,
1932                      $bad_attvals,
1933                      $add_attr_to_tag,
1934                      $message,
1935                      $id,
1936                      $mailbox
1937                      ){
1938     $me = 'sq_sanitize';
1939     $rm_tags = array_shift($tag_list);
1940     /**
1941      * Normalize rm_tags and rm_tags_with_content.
1942      */
1943     @array_walk($tag_list, 'sq_casenormalize');
1944     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1945     @array_walk($self_closing_tags, 'sq_casenormalize');
1946     /**
1947      * See if tag_list is of tags to remove or tags to allow.
1948      * false  means remove these tags
1949      * true   means allow these tags
1950      */
1951     $curpos = 0;
1952     $open_tags = Array();
1953     $trusted = "\n<!-- begin sanitized html -->\n";
1954     $skip_content = false;
1955     /**
1956      * Take care of netscape's stupid javascript entities like
1957      * &{alert('boo')};
1958      */
1959     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1960
1961     while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1962         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1963         $free_content = substr($body, $curpos, $lt-$curpos);
1964         /**
1965          * Take care of <style>
1966          */
1967         if ($tagname == "style" && $tagtype == 1){
1968             list($free_content, $curpos) =
1969                 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
1970             if ($free_content != FALSE){
1971                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1972                 $trusted .= $free_content;
1973                 $trusted .= sq_tagprint($tagname, false, 2);
1974             }
1975             continue;
1976         }
1977         if ($skip_content == false){
1978             $trusted .= $free_content;
1979         }
1980         if ($tagname != FALSE){
1981             if ($tagtype == 2){
1982                 if ($skip_content == $tagname){
1983                     /**
1984                      * Got to the end of tag we needed to remove.
1985                      */
1986                     $tagname = false;
1987                     $skip_content = false;
1988                 } else {
1989                     if ($skip_content == false){
1990                         if ($tagname == "body"){
1991                             $tagname = "div";
1992                         }
1993                         if (isset($open_tags{$tagname}) &&
1994                                 $open_tags{$tagname} > 0){
1995                             $open_tags{$tagname}--;
1996                         } else {
1997                             $tagname = false;
1998                         }
1999                     }
2000                 }
2001             } else {
2002                 /**
2003                  * $rm_tags_with_content
2004                  */
2005                 if ($skip_content == false){
2006                     /**
2007                      * See if this is a self-closing type and change
2008                      * tagtype appropriately.
2009                      */
2010                     if ($tagtype == 1
2011                             && in_array($tagname, $self_closing_tags)){
2012                         $tagtype = 3;
2013                     }
2014                     /**
2015                      * See if we should skip this tag and any content
2016                      * inside it.
2017                      */
2018                     if ($tagtype == 1 &&
2019                             in_array($tagname, $rm_tags_with_content)){
2020                         $skip_content = $tagname;
2021                     } else {
2022                         if (($rm_tags == false
2023                                     && in_array($tagname, $tag_list)) ||
2024                                 ($rm_tags == true &&
2025                                  !in_array($tagname, $tag_list))){
2026                             $tagname = false;
2027                         } else {
2028                             /**
2029                              * Convert body into div.
2030                              */
2031                             if ($tagname == "body"){
2032                                 $tagname = "div";
2033                                 $attary = sq_body2div($attary, $mailbox,
2034                                         $message, $id);
2035                             }
2036                             if ($tagtype == 1){
2037                                 if (isset($open_tags{$tagname})){
2038                                     $open_tags{$tagname}++;
2039                                 } else {
2040                                     $open_tags{$tagname}=1;
2041                                 }
2042                             }
2043                             /**
2044                              * This is where we run other checks.
2045                              */
2046                             if (is_array($attary) && sizeof($attary) > 0){
2047                                 $attary = sq_fixatts($tagname,
2048                                                      $attary,
2049                                                      $rm_attnames,
2050                                                      $bad_attvals,
2051                                                      $add_attr_to_tag,
2052                                                      $message,
2053                                                      $id,
2054                                                      $mailbox
2055                                                      );
2056                             }
2057                         }
2058                     }
2059                 }
2060             }
2061             if ($tagname != false && $skip_content == false){
2062                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2063             }
2064         }
2065         $curpos = $gt+1;
2066     }
2067     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2068     if ($force_tag_closing == true){
2069         foreach ($open_tags as $tagname=>$opentimes){
2070             while ($opentimes > 0){
2071                 $trusted .= '</' . $tagname . '>';
2072                 $opentimes--;
2073             }
2074         }
2075         $trusted .= "\n";
2076     }
2077     $trusted .= "<!-- end sanitized html -->\n";
2078     return $trusted;
2079 }
2080
2081 /**
2082  * This is a wrapper function to call html sanitizing routines.
2083  *
2084  * @param  $body  the body of the message
2085  * @param  $id    the id of the message
2086
2087  * @param  $message
2088  * @param  $mailbox
2089  * @param  boolean $take_mailto_links When TRUE, converts mailto: links
2090  *                                    into internal SM compose links
2091  *                                    (optional; default = TRUE)
2092  * @return        a string with html safe to display in the browser.
2093  */
2094 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2095
2096     // require_once(SM_PATH . 'functions/url_parser.php');  // for $MailTo_PReg_Match
2097
2098     global $attachment_common_show_images, $view_unsafe_images,
2099            $has_unsafe_images;
2100     /**
2101      * Don't display attached images in HTML mode.
2102      */
2103     $attachment_common_show_images = false;
2104     $tag_list = Array(
2105             false,
2106             "object",
2107             "meta",
2108             "html",
2109             "head",
2110             "base",
2111             "link",
2112             "frame",
2113             "iframe",
2114             "plaintext",
2115             "marquee"
2116             );
2117
2118     $rm_tags_with_content = Array(
2119             "script",
2120             "applet",
2121             "embed",
2122             "title",
2123             "frameset",
2124             "xmp",
2125             "xml"
2126             );
2127
2128     $self_closing_tags =  Array(
2129             "img",
2130             "br",
2131             "hr",
2132             "input",
2133             "outbind"
2134             );
2135
2136     $force_tag_closing = true;
2137
2138     $rm_attnames = Array(
2139             "/.*/" =>
2140             Array(
2141                 "/target/i",
2142                 "/^on.*/i",
2143                 "/^dynsrc/i",
2144                 "/^data.*/i",
2145                 "/^lowsrc.*/i"
2146                 )
2147             );
2148
2149     $secremoveimg = "../images/" . _("sec_remove_eng.png");
2150     $bad_attvals = Array(
2151             "/.*/" =>
2152             Array(
2153                 "/^src|background/i" =>
2154                 Array(
2155                     Array(
2156                         "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2157                         "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2158                         "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2159                         ),
2160                     Array(
2161                         "\\1$secremoveimg\\2",
2162                         "\\1$secremoveimg\\2",
2163                         "\\1$secremoveimg\\2",
2164                         )
2165                     ),
2166                 "/^href|action/i" =>
2167                 Array(
2168                     Array(
2169                         "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2170                         "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2171                         "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2172                         ),
2173                     Array(
2174                         "\\1#\\1",
2175                         "\\1#\\1",
2176                         "\\1#\\1"
2177                         )
2178                     ),
2179         "/^style/i" =>
2180             Array(
2181                 Array(
2182                     "/\/\*.*\*\//",
2183                     "/expression/i",
2184                     "/binding/i",
2185                     "/behaviou*r/i",
2186                     "/include-source/i",
2187                     "/position\s*:\s*absolute/i",
2188                     "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
2189                     "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2190                     "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2191                     "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2192                     "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2193                     ),
2194                 Array(
2195                     "",
2196                     "idiocy",
2197                     "idiocy",
2198                     "idiocy",
2199                     "idiocy",
2200                     "",
2201                     "url",
2202                     "url(\\1#\\1)",
2203                     "url(\\1#\\1)",
2204                     "url(\\1#\\1)",
2205                     "\\1:url(\\2#\\3)"
2206                     )
2207                 )
2208             )
2209         );
2210     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
2211         $view_unsafe_images = false;
2212     }
2213     if (!$view_unsafe_images){
2214         /**
2215          * Remove any references to http/https if view_unsafe_images set
2216          * to false.
2217          */
2218         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2219                 '/^([\'\"])\s*https*:.*([\'\"])/si');
2220         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2221                 "\\1$secremoveimg\\1");
2222         array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2223                 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2224         array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2225                 "url(\\1$secremoveimg\\1)");
2226     }
2227
2228     $add_attr_to_tag = Array(
2229             "/^a$/i" =>
2230             Array('target'=>'"_blank"',
2231                 'title'=>'"'._("This external link will open in a new window").'"'
2232                 )
2233             );
2234     $trusted = sq_sanitize($body,
2235                            $tag_list,
2236                            $rm_tags_with_content,
2237                            $self_closing_tags,
2238                            $force_tag_closing,
2239                            $rm_attnames,
2240                            $bad_attvals,
2241                            $add_attr_to_tag,
2242                            $message,
2243                            $id,
2244                            $mailbox
2245                            );
2246     if (preg_match("|$secremoveimg|i", $trusted)){
2247         $has_unsafe_images = true;
2248     }
2249
2250     // we want to parse mailto's in HTML output, change to SM compose links
2251     // this is a modified version of code from url_parser.php... but Marc is
2252     // right: we need a better filtering implementation; adding this randomly
2253     // here is not a great solution
2254     //
2255     if ($take_mailto_links) {
2256         // parseUrl($trusted);   // this even parses URLs inside of tags... too aggressive
2257         global $MailTo_PReg_Match;
2258         $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1) ;
2259         if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2260             foreach ($regs[0] as $i => $mailto_before) {
2261                 $mailto_params = $regs[10][$i];
2262                 // get rid of any tailing quote since we have to add send_to to the end
2263                 //
2264                 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2265                     $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2266                 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2267                     $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2268
2269                 if ($regs[1][$i]) {    //if there is an email addr before '?', we need to merge it with the params
2270                     $to = 'to=' . $regs[1][$i];
2271                     if (strpos($mailto_params, 'to=') > -1)    //already a 'to='
2272                         $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2273                     else {
2274                         if ($mailto_params)    //already some params, append to them
2275                             $mailto_params .= '&amp;' . $to;
2276                         else
2277                             $mailto_params .= '?' . $to;
2278                     }
2279                 }
2280
2281                 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2282
2283                 // we'll already have target=_blank, no need to allow comp_in_new
2284                 // here (which would be a lot more work anyway)
2285                 //
2286                 global $compose_new_win;
2287                 $temp_comp_in_new = $compose_new_win;
2288                 $compose_new_win = 0;
2289                 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2290                 $compose_new_win = $temp_comp_in_new;
2291
2292                 // remove <a href=" and anything after the next quote (we only
2293                 // need the uri, not the link HTML) in compose uri
2294                 //
2295                 $comp_uri = substr($comp_uri, 9);
2296                 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2297                 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2298             }
2299         }
2300     }
2301
2302     return $trusted;
2303 }
2304
2305 /**
2306  * function SendDownloadHeaders - send file to the browser
2307  *
2308  * Original Source: SM core src/download.php
2309  * moved here to make it available to other code, and separate
2310  * front end from back end functionality.
2311  *
2312  * @param string $type0 first half of mime type
2313  * @param string $type1 second half of mime type
2314  * @param string $filename filename to tell the browser for downloaded file
2315  * @param boolean $force whether to force the download dialog to pop
2316  * @param optional integer $filesize send the Content-Header and length to the browser
2317  * @return void
2318  */
2319 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2320     global $languages, $squirrelmail_language;
2321     $isIE = $isIE6plus = false;
2322
2323     sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2324
2325     if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2326             strstr($HTTP_USER_AGENT, 'Opera') === false) {
2327         $isIE = true;
2328     }
2329
2330     if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2331         ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2332         $isIE6plus = true;
2333     }
2334
2335     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2336             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2337         $filename =
2338             call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2339     } else {
2340         $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
2341     }
2342
2343     // A Pox on Microsoft and it's Internet Explorer!
2344     //
2345     // IE has lots of bugs with file downloads.
2346     // It also has problems with SSL.  Both of these cause problems
2347     // for us in this function.
2348     //
2349     // See this article on Cache Control headers and SSL
2350     // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2351     //
2352     // The best thing you can do for IE is to upgrade to the latest
2353     // version
2354     //set all the Cache Control Headers for IE
2355     if ($isIE) {
2356         $filename=rawurlencode($filename);
2357         header ("Pragma: public");
2358         header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2359         header ("Cache-Control: post-check=0, pre-check=0", false);
2360         header ("Cache-Control: private");
2361
2362         //set the inline header for IE, we'll add the attachment header later if we need it
2363         header ("Content-Disposition: inline; filename=$filename");
2364     }
2365
2366     if (!$force) {
2367         // Try to show in browser window
2368         header ("Content-Disposition: inline; filename=\"$filename\"");
2369         header ("Content-Type: $type0/$type1; name=\"$filename\"");
2370     } else {
2371         // Try to pop up the "save as" box
2372
2373         // IE makes this hard.  It pops up 2 save boxes, or none.
2374         // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2375         // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2376         // But, according to Microsoft, it is "RFC compliant but doesn't
2377         // take into account some deviations that allowed within the
2378         // specification."  Doesn't that mean RFC non-compliant?
2379         // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2380
2381         // all browsers need the application/octet-stream header for this
2382         header ("Content-Type: application/octet-stream; name=\"$filename\"");
2383
2384         // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2385         // Do not have quotes around filename, but that applied to
2386         // "attachment"... does it apply to inline too?
2387         header ("Content-Disposition: attachment; filename=\"$filename\"");
2388
2389         if ($isIE && !$isIE6plus) {
2390             // This combination seems to work mostly.  IE 5.5 SP 1 has
2391             // known issues (see the Microsoft Knowledge Base)
2392
2393             // This works for most types, but doesn't work with Word files
2394             header ("Content-Type: application/download; name=\"$filename\"");
2395
2396             // These are spares, just in case.  :-)
2397             //header("Content-Type: $type0/$type1; name=\"$filename\"");
2398             //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2399             //header("Content-Type: application/octet-stream; name=\"$filename\"");
2400         } else {
2401             // another application/octet-stream forces download for Netscape
2402             header ("Content-Type: application/octet-stream; name=\"$filename\"");
2403         }
2404     }
2405
2406     //send the content-length header if the calling function provides it
2407     if ($filesize > 0) {
2408         header("Content-Length: $filesize");
2409     }
2410
2411 }  // end fn SendDownloadHeaders