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            $download_href, $unsafe_image_toggle_href, $unsafe_image_toggle_text;
 360
 361     // workaround for not updated config.php
 362     if (! isset($use_iframe)) $use_iframe = false;
 363
 364     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
 365         $view_unsafe_images = false;
 366     }
 367
 368     $body = '';
 369     $urlmailbox = urlencode($mailbox);
 370     $body_message = getEntity($message, $ent_num);
 371     if (($body_message->header->type0 == 'text') ||
 372             ($body_message->header->type0 == 'rfc822')) {
 373         $body = mime_fetch_body ($imap_stream, $id, $ent_num);
 374         $body = decodeBody($body, $body_message->header->encoding);
 375
 376         if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 377                 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
 378             if (mb_detect_encoding($body) != 'ASCII') {
 379                 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
 380             }
 381         }
 382         $hookResults = do_hook("message_body", $body);
 383         $body = $hookResults[1];
 384
 385         /* If there are other types that shouldn't be formatted, add
 386          * them here.
 387          */
 388
 389         if ($body_message->header->type1 == 'html') {
 390             if ($show_html_default <> 1) {
 391                 $entity_conv = array('&nbsp;' => ' ',
 392                                      '<p>'    => "\n",
 393                                      '<P>'    => "\n",
 394                                      '<br>'   => "\n",
 395                                      '<BR>'   => "\n",
 396                                      '<br />' => "\n",
 397                                      '<BR />' => "\n",
 398                                      '&gt;'   => '>',
 399                                      '&lt;'   => '<');
 400                 $body = strtr($body, $entity_conv);
 401                 $body = strip_tags($body);
 402                 $body = trim($body);
 403                 translateText($body, $wrap_at,
 404                         $body_message->header->getParameter('charset'));
 405             } elseif ($use_iframe && ! $clean) {
 406                 // $clean is used to remove iframe in printable view.
 407
 408                 /**
 409                  * If we don't add html message between iframe tags,
 410                  * we must detect unsafe images and modify $has_unsafe_images.
 411                  */
 412                 $html_body = magicHTML($body, $id, $message, $mailbox);
 413                 // Convert character set in order to display html mails in different character set
 414                 $html_body = charset_decode($body_message->header->getParameter('charset'),$html_body,false,true);
 415
 416                 // creating iframe url
 417                 $iframeurl=sqm_baseuri().'src/view_html.php?'
 418                     . 'mailbox=' . $urlmailbox
 419                     . '&amp;passed_id=' . $id
 420                     . '&amp;ent_id=' . $ent_num
 421                     . '&amp;view_unsafe_images=' . (int) $view_unsafe_images;
 422
 423                 // adding warning message
 424                 $body = html_tag('div',_("Viewing HTML formatted email"),'center');
 425
 426                 /**
 427                  * height can't be set to 100%, because it does not work as expected when
 428                  * iframe is inside the table. Browsers do not create full height objects
 429                  * even when iframe is not nested. Maybe there is some way to get full size
 430                  * with CSS. Tested in firefox 1.02 and opera 7.53
 431                  *
 432                  * width="100%" does not work as expected, when table width is not set (automatic)
 433                  *
 434                  * tokul: I think <iframe> are safer sandbox than <object>. Objects might
 435                  * need special handling for IE and IE6SP2.
 436                  */
 437                 $body.= "<div><iframe name=\"message_frame\" width=\"100%\" height=\"$iframe_height\" src=\"$iframeurl\""
 438                     .' frameborder="1" marginwidth="0" marginheight="0" scrolling="auto">' . "\n";
 439
 440                 // Message for browsers without iframe support
 441                 //$body.= _("Your browser does not support inline frames.
 442                 // You can view HTML formated message by following below link.");
 443                 //$body.= "<br /><a href=\"$iframeurl\">"._("View HTML Message")."</a>";
 444
 445                 // if browser can't render iframe, it renders html message.
 446                 $body.= $html_body;
 447
 448                 // close iframe
 449                 $body.="</iframe></div>\n";
 450             } else {
 451                 // old way of html rendering
 452                 $body = magicHTML($body, $id, $message, $mailbox);
 453                 /**
 454                  * convert character set. charset_decode does not remove html special chars
 455                  * applied by magicHTML functions and does not sanitize them second time if
 456                  * fourth argument is true.
 457                  */
 458                 $body = charset_decode($body_message->header->getParameter('charset'),$body,false,true);
 459             }
 460         } else {
 461             translateText($body, $wrap_at,
 462                     $body_message->header->getParameter('charset'));
 463         }
 464
 465         // if this is the clean display (i.e. printer friendly), stop here.
 466         if ( $clean ) {
 467             return $body;
 468         }
 469
 470         $download_and_unsafe_link = '';
 471
 472         $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
 473             '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
 474             '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 475         if (isset($passed_ent_id)) {
 476             $link .= '&amp;passed_ent_id='.$passed_ent_id;
 477         }
 478         $download_href = SM_PATH . 'src/download.php?absolute_dl=true&amp;' . $link;
 479         $download_and_unsafe_link .= '&nbsp;|&nbsp;<a href="'. $download_href .'">' . _("Download this as a file") .  '</a>';
 480         if ($view_unsafe_images) {
 481             $text = _("Hide Unsafe Images");
 482         } else {
 483             if (isset($has_unsafe_images) && $has_unsafe_images) {
 484                 $link .= '&amp;view_unsafe_images=1';
 485                 $text = _("View Unsafe Images");
 486             } else {
 487                 $text = '';
 488             }
 489         }
 490         if($text != '') {
 491             $unsafe_image_toggle_href = SM_PATH . 'src/read_body.php?'.$link;
 492             $unsafe_image_toggle_text = $text;
 493             $download_and_unsafe_link .= '&nbsp;|&nbsp;<a href="'. $unsafe_image_toggle_href .'">' . $text . '</a>';
 494         }
 495     }
 496     return $body;
 497 }
 498
 499 /**
 500  * Displays attachment links and information
 501  *
 502  * Since 1.3.0 function is not included in formatBody() call.
 503  *
 504  * Since 1.0.2 uses attachment $type0/$type1 hook.
 505  * Since 1.2.5 uses attachment $type0/* hook.
 506  * Since 1.5.0 uses attachments_bottom hook.
 507  *
 508  * @param object $message SquirrelMail message object
 509  * @param array $exclude_id message parts that are not attachments.
 510  * @param string $mailbox mailbox name
 511  * @param integer $id message id
 512  * @return string html formated attachment information.
 513  */
 514 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 515     global $where, $what, $startMessage, $color, $passed_ent_id, $base_uri;
 516
 517     $att_ar = $message->getAttachments($exclude_id);
 518
 519     if (!count($att_ar)) return '';
 520
 521     $attachments = '';
 522
 523     $urlMailbox = urlencode($mailbox);
 524
 525     foreach ($att_ar as $att) {
 526         $ent = $att->entity_id;
 527         $header = $att->header;
 528         $type0 = strtolower($header->type0);
 529         $type1 = strtolower($header->type1);
 530         $name = '';
 531         $links['download link']['text'] = _("Download");
 532         $links['download link']['href'] = $base_uri .
 533             "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 534         if ($type0 =='message' && $type1 == 'rfc822') {
 535             $default_page = $base_uri  . 'src/read_body.php';
 536             $rfc822_header = $att->rfc822_header;
 537             $filename = $rfc822_header->subject;
 538             if (trim( $filename ) == '') {
 539                 $filename = 'untitled-[' . $ent . ']' ;
 540             }
 541             $from_o = $rfc822_header->from;
 542             if (is_object($from_o)) {
 543                 $from_name = decodeHeader($from_o->getAddress(false));
 544             } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
 545                 // something weird happens when a digest message is opened and you return to the digest
 546                 // now the from object is part of an array. Probably the parseHeader call overwrites the info
 547                 // retrieved from the bodystructure in a different way. We need to fix this later.
 548                 // possible starting point, do not fetch header we already have and inspect how
 549                 // the rfc822_header object behaves.
 550                 $from_name = decodeHeader($from_o[0]->getAddress(false));
 551             } else {
 552                 $from_name = _("Unknown sender");
 553             }
 554             $description = $from_name;
 555         } else {
 556             $default_page = $base_uri  . 'src/download.php';
 557             $filename = $att->getFilename();
 558             if ($header->description) {
 559                 $description = decodeHeader($header->description);
 560             } else {
 561                 $description = '';
 562             }
 563         }
 564
 565         $display_filename = $filename;
 566         if (isset($passed_ent_id)) {
 567             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 568         } else {
 569             $passed_ent_id_link = '';
 570         }
 571         $defaultlink = $default_page . "?startMessage=$startMessage"
 572             . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 573             . '&amp;ent_id='.$ent.$passed_ent_id_link;
 574         if ($where && $what) {
 575             $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 576         }
 577
 578         /* This executes the attachment hook with a specific MIME-type.
 579          * If that doesn't have results, it tries if there's a rule
 580          * for a more generic type.
 581          */
 582         $hookresults = do_hook("attachment $type0/$type1", $links,
 583                 $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 584                 $display_filename, $where, $what);
 585         if(count($hookresults[1]) <= 1) {
 586             $hookresults = do_hook("attachment $type0/*", $links,
 587                     $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 588                     $display_filename, $where, $what);
 589         }
 590
 591         $links = $hookresults[1];
 592         $defaultlink = $hookresults[6];
 593
 594         $attachments .= '<tr><td>' .
 595             '<a href="'.$defaultlink.'">'.decodeHeader($display_filename).'</a>&nbsp;</td>' .
 596             '<td><small><b>' . show_readable_size($header->size) .
 597             '</b>&nbsp;&nbsp;</small></td>' .
 598             '<td><small>[ '.htmlspecialchars($type0).'/'.htmlspecialchars($type1).' ]&nbsp;</small></td>' .
 599             '<td><small>';
 600         $attachments .= '<b>' . $description . '</b>';
 601         $attachments .= '</small></td><td><small>&nbsp;';
 602
 603         $skipspaces = 1;
 604         foreach ($links as $val) {
 605             if ($skipspaces) {
 606                 $skipspaces = 0;
 607             } else {
 608                 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
 609             }
 610             $attachments .= '<a href="' . $val['href'] . '">'
 611                 . (isset($val['text']) && !empty($val['text']) ? $val['text'] : '')
 612                 . (isset($val['extra']) && !empty($val['extra']) ? $val['extra'] : '')
 613                 . '</a>';
 614         }
 615         unset($links);
 616         $attachments .= "</td></tr>\n";
 617     }
 618     $attachmentadd = do_hook_function('attachments_bottom',$attachments);
 619     if ($attachmentadd != '')
 620         $attachments = $attachmentadd;
 621     return $attachments;
 622 }
 623
 624 function sqimap_base64_decode(&$string) {
 625
 626     // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
 627     // fly decoding (to reduce memory usage) you have to check if the
 628     // data has incomplete pairs
 629
 630     // Remove the noise in order to check if the 4 bytes pairs are complete
 631     $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
 632
 633     $sStringRem = '';
 634     $iMod = strlen($string) % 4;
 635     if ($iMod) {
 636         $sStringRem = substr($string,-$iMod);
 637         // Check if $sStringRem contains padding characters
 638         if (substr($sStringRem,-1) != '=') {
 639             $string = substr($string,0,-$iMod);
 640         } else {
 641             $sStringRem = '';
 642         }
 643     }
 644     $string = base64_decode($string);
 645     return $sStringRem;
 646 }
 647
 648 /**
 649  * Decodes encoded message body
 650  *
 651  * This function decodes the body depending on the encoding type.
 652  * Currently quoted-printable and base64 encodings are supported.
 653  * decode_body hook was added to this function in 1.4.2/1.5.0
 654  * @param string $body encoded message body
 655  * @param string $encoding used encoding
 656  * @return string decoded string
 657  * @since 1.0
 658  */
 659 function decodeBody($body, $encoding) {
 660
 661     $body = str_replace("\r\n", "\n", $body);
 662     $encoding = strtolower($encoding);
 663
 664     $encoding_handler = do_hook_function('decode_body', $encoding);
 665
 666
 667     // plugins get first shot at decoding the body
 668     //
 669     if (!empty($encoding_handler) && function_exists($encoding_handler)) {
 670         $body = $encoding_handler('decode', $body);
 671
 672     } elseif ($encoding == 'quoted-printable' ||
 673             $encoding == 'quoted_printable') {
 674         /**
 675          * quoted_printable_decode() function is broken in older
 676          * php versions. Text with \r\n decoding was fixed only
 677          * in php 4.3.0. Minimal code requirement 4.0.4 +
 678          * str_replace("\r\n", "\n", $body); call.
 679          */
 680         $body = quoted_printable_decode($body);
 681     } elseif ($encoding == 'base64') {
 682         $body = base64_decode($body);
 683     }
 684
 685     // All other encodings are returned raw.
 686     return $body;
 687 }
 688
 689 /**
 690  * Decodes headers
 691  *
 692  * This functions decode strings that is encoded according to
 693  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 694  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 695  *
 696  * @param string $string header string that has to be made readable
 697  * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
 698  * @param boolean $htmlsave preserve spaces and sanitize html special characters. defaults to true
 699  * @param boolean $decide decide if string can be utfencoded. defaults to false
 700  * @return string decoded header string
 701  */
 702 function decodeHeader ($string, $utfencode=true,$htmlsave=true,$decide=false) {
 703     global $languages, $squirrelmail_language,$default_charset;
 704     if (is_array($string)) {
 705         $string = implode("\n", $string);
 706     }
 707
 708     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 709             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
 710         $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
 711         // Do we need to return at this point?
 712         // return $string;
 713     }
 714     $i = 0;
 715     $iLastMatch = -2;
 716     $encoded = true;
 717
 718     $aString = explode(' ',$string);
 719     $ret = '';
 720     foreach ($aString as $chunk) {
 721         if ($encoded && $chunk === '') {
 722             continue;
 723         } elseif ($chunk === '') {
 724             $ret .= ' ';
 725             continue;
 726         }
 727         $encoded = false;
 728         /* if encoded words are not separated by a linear-space-white we still catch them */
 729         $j = $i-1;
 730
 731         while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 732             /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
 733             if ($iLastMatch !== $j) {
 734                 if ($htmlsave) {
 735                     $ret .= '&#32;';
 736                 } else {
 737                     $ret .= ' ';
 738                 }
 739             }
 740             $iLastMatch = $i;
 741             $j = $i;
 742             if ($htmlsave) {
 743                 $ret .= htmlspecialchars($res[1]);
 744             } else {
 745                 $ret .= $res[1];
 746             }
 747             $encoding = ucfirst($res[3]);
 748
 749             /* decide about valid decoding */
 750             if ($decide && is_conversion_safe($res[2])) {
 751                 $utfencode=true;
 752                 $can_be_encoded=true;
 753             } else {
 754                 $can_be_encoded=false;
 755             }
 756             switch ($encoding)
 757             {
 758                 case 'B':
 759                     $replace = base64_decode($res[4]);
 760                     if ($utfencode) {
 761                         if ($can_be_encoded) {
 762                             /* convert string to different charset,
 763                              * if functions asks for it (usually in compose)
 764                              */
 765                             $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsave);
 766                         } else {
 767                             // convert string to html codes in order to display it
 768                             $ret .= charset_decode($res[2],$replace);
 769                         }
 770                     } else {
 771                         if ($htmlsave) {
 772                             $replace = htmlspecialchars($replace);
 773                         }
 774                         $ret.= $replace;
 775                     }
 776                     break;
 777                 case 'Q':
 778                     $replace = str_replace('_', ' ', $res[4]);
 779                     $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 780                             $replace);
 781                     if ($utfencode) {
 782                         if ($can_be_encoded) {
 783                             /* convert string to different charset,
 784                              * if functions asks for it (usually in compose)
 785                              */
 786                             $replace = charset_convert($res[2], $replace,$default_charset,$htmlsave);
 787                         } else {
 788                             // convert string to html codes in order to display it
 789                             $replace = charset_decode($res[2], $replace);
 790                         }
 791                     } else {
 792                         if ($htmlsave) {
 793                             $replace = htmlspecialchars($replace);
 794                         }
 795                     }
 796                     $ret .= $replace;
 797                     break;
 798                 default:
 799                     break;
 800             }
 801             $chunk = $res[5];
 802             $encoded = true;
 803         }
 804         if (!$encoded) {
 805             if ($htmlsave) {
 806                 $ret .= '&#32;';
 807             } else {
 808                 $ret .= ' ';
 809             }
 810         }
 811
 812         if (!$encoded && $htmlsave) {
 813             $ret .= htmlspecialchars($chunk);
 814         } else {
 815             $ret .= $chunk;
 816         }
 817         ++$i;
 818     }
 819     /* remove the first added space */
 820     if ($ret) {
 821         if ($htmlsave) {
 822             $ret = substr($ret,5);
 823         } else {
 824             $ret = substr($ret,1);
 825         }
 826     }
 827
 828     return $ret;
 829 }
 830
 831 /**
 832  * Encodes header
 833  *
 834  * Function uses XTRA_CODE _encodeheader function, if such function exists.
 835  *
 836  * Function uses Q encoding by default and encodes a string according to RFC
 837  * 1522 for use in headers if it contains 8-bit characters or anything that
 838  * looks like it should be encoded.
 839  *
 840  * Function switches to B encoding and encodeHeaderBase64() function, if
 841  * string is 8bit and multibyte character set supported by mbstring extension
 842  * is used. It can cause E_USER_NOTICE errors, if interface is used with
 843  * multibyte character set unsupported by mbstring extension.
 844  *
 845  * @param string $string header string, that has to be encoded
 846  * @return string quoted-printable encoded string
 847  * @todo make $mb_charsets system wide constant
 848  */
 849 function encodeHeader ($string) {
 850     global $default_charset, $languages, $squirrelmail_language;
 851
 852     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 853             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
 854         return  call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
 855     }
 856
 857     // Use B encoding for multibyte charsets
 858     $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
 859     if (in_array($default_charset,$mb_charsets) &&
 860         in_array($default_charset,sq_mb_list_encodings()) &&
 861         sq_is8bit($string)) {
 862         return encodeHeaderBase64($string,$default_charset);
 863     } elseif (in_array($default_charset,$mb_charsets) &&
 864               sq_is8bit($string) &&
 865               ! in_array($default_charset,sq_mb_list_encodings())) {
 866         // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
 867         // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
 868     }
 869
 870     // Encode only if the string contains 8-bit characters or =?
 871     $j = strlen($string);
 872     $max_l = 75 - strlen($default_charset) - 7;
 873     $aRet = array();
 874     $ret = '';
 875     $iEncStart = $enc_init = false;
 876     $cur_l = $iOffset = 0;
 877     for($i = 0; $i < $j; ++$i) {
 878         switch($string{$i})
 879         {
 880             case '=':
 881             case '<':
 882             case '>':
 883             case ',':
 884             case '?':
 885             case '_':
 886                 if ($iEncStart === false) {
 887                     $iEncStart = $i;
 888                 }
 889                 $cur_l+=3;
 890                 if ($cur_l > ($max_l-2)) {
 891                     /* if there is an stringpart that doesn't need encoding, add it */
 892                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 893                     $aRet[] = "=?$default_charset?Q?$ret?=";
 894                     $iOffset = $i;
 895                     $cur_l = 0;
 896                     $ret = '';
 897                     $iEncStart = false;
 898                 } else {
 899                     $ret .= sprintf("=%02X",ord($string{$i}));
 900                 }
 901                 break;
 902             case '(':
 903             case ')':
 904                 if ($iEncStart !== false) {
 905                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 906                     $aRet[] = "=?$default_charset?Q?$ret?=";
 907                     $iOffset = $i;
 908                     $cur_l = 0;
 909                     $ret = '';
 910                     $iEncStart = false;
 911                 }
 912                 break;
 913             case ' ':
 914                 if ($iEncStart !== false) {
 915                     $cur_l++;
 916                     if ($cur_l > $max_l) {
 917                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 918                         $aRet[] = "=?$default_charset?Q?$ret?=";
 919                         $iOffset = $i;
 920                         $cur_l = 0;
 921                         $ret = '';
 922                         $iEncStart = false;
 923                     } else {
 924                         $ret .= '_';
 925                     }
 926                 }
 927                 break;
 928             default:
 929                 $k = ord($string{$i});
 930                 if ($k > 126) {
 931                     if ($iEncStart === false) {
 932                         // do not start encoding in the middle of a string, also take the rest of the word.
 933                         $sLeadString = substr($string,0,$i);
 934                         $aLeadString = explode(' ',$sLeadString);
 935                         $sToBeEncoded = array_pop($aLeadString);
 936                         $iEncStart = $i - strlen($sToBeEncoded);
 937                         $ret .= $sToBeEncoded;
 938                         $cur_l += strlen($sToBeEncoded);
 939                     }
 940                     $cur_l += 3;
 941                     /* first we add the encoded string that reached it's max size */
 942                     if ($cur_l > ($max_l-2)) {
 943                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 944                         $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
 945                         $cur_l = 3;
 946                         $ret = '';
 947                         $iOffset = $i;
 948                         $iEncStart = $i;
 949                     }
 950                     $enc_init = true;
 951                     $ret .= sprintf("=%02X", $k);
 952                 } else {
 953                     if ($iEncStart !== false) {
 954                         $cur_l++;
 955                         if ($cur_l > $max_l) {
 956                             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 957                             $aRet[] = "=?$default_charset?Q?$ret?=";
 958                             $iEncStart = false;
 959                             $iOffset = $i;
 960                             $cur_l = 0;
 961                             $ret = '';
 962                         } else {
 963                             $ret .= $string{$i};
 964                         }
 965                     }
 966                 }
 967                 break;
 968         }
 969     }
 970
 971     if ($enc_init) {
 972         if ($iEncStart !== false) {
 973             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 974             $aRet[] = "=?$default_charset?Q?$ret?=";
 975         } else {
 976             $aRet[] = substr($string,$iOffset);
 977         }
 978         $string = implode('',$aRet);
 979     }
 980     return $string;
 981 }
 982
 983 /**
 984  * Encodes string according to rfc2047 B encoding header formating rules
 985  *
 986  * It is recommended way to encode headers with character sets that store
 987  * symbols in more than one byte.
 988  *
 989  * Function requires mbstring support. If required mbstring functions are missing,
 990  * function returns false and sets E_USER_WARNING level error message.
 991  *
 992  * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
 993  * that mbstring functions will generate E_WARNING errors, if unsupported
 994  * character set is used. mb_encode_mimeheader function provided by php
 995  * mbstring extension is not used in order to get better control of header
 996  * encoding.
 997  *
 998  * Used php code functions - function_exists(), trigger_error(), strlen()
 999  * (is used with charset names and base64 strings). Used php mbstring
1000  * functions - mb_strlen and mb_substr.
1001  *
1002  * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
1003  * encoding), rfc 2822 (header folding)
1004  *
1005  * @param string $string header string that must be encoded
1006  * @param string $charset character set. Must be supported by mbstring extension.
1007  * Use sq_mb_list_encodings() to detect supported charsets.
1008  * @return string string encoded according to rfc2047 B encoding formating rules
1009  * @since 1.5.1
1010  * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1011  * @todo Do we want to control max length of header?
1012  * @todo Do we want to control EOL (end-of-line) marker?
1013  * @todo Do we want to translate error message?
1014  */
1015 function encodeHeaderBase64($string,$charset) {
1016     /**
1017      * Check mbstring function requirements.
1018      */
1019     if (! function_exists('mb_strlen') ||
1020         ! function_exists('mb_substr')) {
1021         // set E_USER_WARNING
1022         trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
1023         // return false
1024         return false;
1025     }
1026
1027     // initial return array
1028     $aRet = array();
1029
1030     /**
1031      * header length = 75 symbols max (same as in encodeHeader)
1032      * remove $charset length
1033      * remove =? ? ?= (5 chars)
1034      * remove 2 more chars (\r\n ?)
1035      */
1036     $iMaxLength = 75 - strlen($charset) - 7;
1037
1038     // set first character position
1039     $iStartCharNum = 0;
1040
1041     // loop through all characters. count characters and not bytes.
1042     for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1043         // encode string from starting character to current character.
1044         $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1045
1046         // Check encoded string length
1047         if(strlen($encoded_string)>$iMaxLength) {
1048             // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1049             $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1050
1051             // set new starting character
1052             $iStartCharNum = $iCharNum-1;
1053
1054             // encode last char (in case it is last character in string)
1055             $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1056         } // if string is shorter than max length - add next character
1057     }
1058
1059     // add last encoded string to array
1060     $aRet[] = $encoded_string;
1061
1062     // set initial return string
1063     $sRet = '';
1064
1065     // loop through encoded strings
1066     foreach($aRet as $string) {
1067         // TODO: Do we want to control EOL (end-of-line) marker
1068         if ($sRet!='') $sRet.= " ";
1069
1070         // add header tags and encoded string to return string
1071         $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1072     }
1073
1074     return $sRet;
1075 }
1076
1077 /* This function trys to locate the entity_id of a specific mime element */
1078 function find_ent_id($id, $message) {
1079     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1080         if ($message->entities[$i]->header->type0 == 'multipart')  {
1081             $ret = find_ent_id($id, $message->entities[$i]);
1082         } else {
1083             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
1084 //                if (sq_check_save_extension($message->entities[$i])) {
1085                 return $message->entities[$i]->entity_id;
1086 //                }
1087             } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1088                 /**
1089                  * This is part of a fix for Outlook Express 6.x generating
1090                  * cid URLs without creating content-id headers
1091                  * @@JA - 20050207
1092                  */
1093                 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1094                     return $message->entities[$i]->entity_id;
1095                 }
1096             }
1097         }
1098     }
1099     return $ret;
1100 }
1101
1102 function sq_check_save_extension($message) {
1103     $filename = $message->getFilename();
1104     $ext = substr($filename, strrpos($filename,'.')+1);
1105     $save_extensions = array('jpg','jpeg','gif','png','bmp');
1106     return in_array($ext, $save_extensions);
1107 }
1108
1109
1110 /**
1111  ** HTMLFILTER ROUTINES
1112  */
1113
1114 /**
1115  * This function checks attribute values for entity-encoded values
1116  * and returns them translated into 8-bit strings so we can run
1117  * checks on them.
1118  *
1119  * @param  $attvalue A string to run entity check against.
1120  * @return           Nothing, modifies a reference value.
1121  */
1122 function sq_defang(&$attvalue){
1123     $me = 'sq_defang';
1124     /**
1125      * Skip this if there aren't ampersands or backslashes.
1126      */
1127     if (strpos($attvalue, '&') === false
1128         && strpos($attvalue, '\\') === false){
1129         return;
1130     }
1131     $m = false;
1132     do {
1133         $m = false;
1134         $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1135         $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1136         $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1137     } while ($m == true);
1138     $attvalue = stripslashes($attvalue);
1139 }
1140
1141 /**
1142  * Kill any tabs, newlines, or carriage returns. Our friends the
1143  * makers of the browser with 95% market value decided that it'd
1144  * be funny to make "java[tab]script" be just as good as "javascript".
1145  *
1146  * @param  attvalue  The attribute value before extraneous spaces removed.
1147  * @return attvalue  Nothing, modifies a reference value.
1148  */
1149 function sq_unspace(&$attvalue){
1150     $me = 'sq_unspace';
1151     if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1152         $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1153                                 Array('',   '',   '',   '',   ''), $attvalue);
1154     }
1155 }
1156
1157 /**
1158  * This function returns the final tag out of the tag name, an array
1159  * of attributes, and the type of the tag. This function is called by
1160  * sq_sanitize internally.
1161  *
1162  * @param  $tagname  the name of the tag.
1163  * @param  $attary   the array of attributes and their values
1164  * @param  $tagtype  The type of the tag (see in comments).
1165  * @return           a string with the final tag representation.
1166  */
1167 function sq_tagprint($tagname, $attary, $tagtype){
1168     $me = 'sq_tagprint';
1169
1170     if ($tagtype == 2){
1171         $fulltag = '</' . $tagname . '>';
1172     } else {
1173         $fulltag = '<' . $tagname;
1174         if (is_array($attary) && sizeof($attary)){
1175             $atts = Array();
1176             while (list($attname, $attvalue) = each($attary)){
1177                 array_push($atts, "$attname=$attvalue");
1178             }
1179             $fulltag .= ' ' . join(" ", $atts);
1180         }
1181         if ($tagtype == 3){
1182             $fulltag .= ' /';
1183         }
1184         $fulltag .= '>';
1185     }
1186     return $fulltag;
1187 }
1188
1189 /**
1190  * A small helper function to use with array_walk. Modifies a by-ref
1191  * value and makes it lowercase.
1192  *
1193  * @param  $val a value passed by-ref.
1194  * @return      void since it modifies a by-ref value.
1195  */
1196 function sq_casenormalize(&$val){
1197     $val = strtolower($val);
1198 }
1199
1200 /**
1201  * This function skips any whitespace from the current position within
1202  * a string and to the next non-whitespace value.
1203  *
1204  * @param  $body   the string
1205  * @param  $offset the offset within the string where we should start
1206  *                 looking for the next non-whitespace character.
1207  * @return         the location within the $body where the next
1208  *                 non-whitespace char is located.
1209  */
1210 function sq_skipspace($body, $offset){
1211     $me = 'sq_skipspace';
1212     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1213     if (sizeof($matches{1})){
1214         $count = strlen($matches{1});
1215         $offset += $count;
1216     }
1217     return $offset;
1218 }
1219
1220 /**
1221  * This function looks for the next character within a string.  It's
1222  * really just a glorified "strpos", except it catches if failures
1223  * nicely.
1224  *
1225  * @param  $body   The string to look for needle in.
1226  * @param  $offset Start looking from this position.
1227  * @param  $needle The character/string to look for.
1228  * @return         location of the next occurance of the needle, or
1229  *                 strlen($body) if needle wasn't found.
1230  */
1231 function sq_findnxstr($body, $offset, $needle){
1232     $me  = 'sq_findnxstr';
1233     $pos = strpos($body, $needle, $offset);
1234     if ($pos === FALSE){
1235         $pos = strlen($body);
1236     }
1237     return $pos;
1238 }
1239
1240 /**
1241  * This function takes a PCRE-style regexp and tries to match it
1242  * within the string.
1243  *
1244  * @param  $body   The string to look for needle in.
1245  * @param  $offset Start looking from here.
1246  * @param  $reg    A PCRE-style regex to match.
1247  * @return         Returns a false if no matches found, or an array
1248  *                 with the following members:
1249  *                 - integer with the location of the match within $body
1250  *                 - string with whatever content between offset and the match
1251  *                 - string with whatever it is we matched
1252  */
1253 function sq_findnxreg($body, $offset, $reg){
1254     $me = 'sq_findnxreg';
1255     $matches = Array();
1256     $retarr = Array();
1257     preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1258     if (!isset($matches{0}) || !$matches{0}){
1259         $retarr = false;
1260     } else {
1261         $retarr{0} = $offset + strlen($matches{1});
1262         $retarr{1} = $matches{1};
1263         $retarr{2} = $matches{2};
1264     }
1265     return $retarr;
1266 }
1267
1268 /**
1269  * This function looks for the next tag.
1270  *
1271  * @param  $body   String where to look for the next tag.
1272  * @param  $offset Start looking from here.
1273  * @return         false if no more tags exist in the body, or
1274  *                 an array with the following members:
1275  *                 - string with the name of the tag
1276  *                 - array with attributes and their values
1277  *                 - integer with tag type (1, 2, or 3)
1278  *                 - integer where the tag starts (starting "<")
1279  *                 - integer where the tag ends (ending ">")
1280  *                 first three members will be false, if the tag is invalid.
1281  */
1282 function sq_getnxtag($body, $offset){
1283     $me = 'sq_getnxtag';
1284     if ($offset > strlen($body)){
1285         return false;
1286     }
1287     $lt = sq_findnxstr($body, $offset, "<");
1288     if ($lt == strlen($body)){
1289         return false;
1290     }
1291     /**
1292      * We are here:
1293      * blah blah <tag attribute="value">
1294      * \---------^
1295      */
1296     $pos = sq_skipspace($body, $lt+1);
1297     if ($pos >= strlen($body)){
1298         return Array(false, false, false, $lt, strlen($body));
1299     }
1300     /**
1301      * There are 3 kinds of tags:
1302      * 1. Opening tag, e.g.:
1303      *    <a href="blah">
1304      * 2. Closing tag, e.g.:
1305      *    </a>
1306      * 3. XHTML-style content-less tag, e.g.:
1307      *    <img src="blah" />
1308      */
1309     $tagtype = false;
1310     switch (substr($body, $pos, 1)){
1311         case '/':
1312             $tagtype = 2;
1313             $pos++;
1314             break;
1315         case '!':
1316             /**
1317              * A comment or an SGML declaration.
1318              */
1319             if (substr($body, $pos+1, 2) == "--"){
1320                 $gt = strpos($body, "-->", $pos);
1321                 if ($gt === false){
1322                     $gt = strlen($body);
1323                 } else {
1324                     $gt += 2;
1325                 }
1326                 return Array(false, false, false, $lt, $gt);
1327             } else {
1328                 $gt = sq_findnxstr($body, $pos, ">");
1329                 return Array(false, false, false, $lt, $gt);
1330             }
1331             break;
1332         default:
1333             /**
1334              * Assume tagtype 1 for now. If it's type 3, we'll switch values
1335              * later.
1336              */
1337             $tagtype = 1;
1338             break;
1339     }
1340
1341     $tag_start = $pos;
1342     $tagname = '';
1343     /**
1344      * Look for next [\W-_], which will indicate the end of the tag name.
1345      */
1346     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1347     if ($regary == false){
1348         return Array(false, false, false, $lt, strlen($body));
1349     }
1350     list($pos, $tagname, $match) = $regary;
1351     $tagname = strtolower($tagname);
1352
1353     /**
1354      * $match can be either of these:
1355      * '>'  indicating the end of the tag entirely.
1356      * '\s' indicating the end of the tag name.
1357      * '/'  indicating that this is type-3 xhtml tag.
1358      *
1359      * Whatever else we find there indicates an invalid tag.
1360      */
1361     switch ($match){
1362         case '/':
1363             /**
1364              * This is an xhtml-style tag with a closing / at the
1365              * end, like so: <img src="blah" />. Check if it's followed
1366              * by the closing bracket. If not, then this tag is invalid
1367              */
1368             if (substr($body, $pos, 2) == "/>"){
1369                 $pos++;
1370                 $tagtype = 3;
1371             } else {
1372                 $gt = sq_findnxstr($body, $pos, ">");
1373                 $retary = Array(false, false, false, $lt, $gt);
1374                 return $retary;
1375             }
1376         case '>':
1377             return Array($tagname, false, $tagtype, $lt, $pos);
1378             break;
1379         default:
1380             /**
1381              * Check if it's whitespace
1382              */
1383             if (!preg_match('/\s/', $match)){
1384                 /**
1385                  * This is an invalid tag! Look for the next closing ">".
1386                  */
1387                 $gt = sq_findnxstr($body, $lt, ">");
1388                 return Array(false, false, false, $lt, $gt);
1389             }
1390             break;
1391     }
1392
1393     /**
1394      * At this point we're here:
1395      * <tagname  attribute='blah'>
1396      * \-------^
1397      *
1398      * At this point we loop in order to find all attributes.
1399      */
1400     $attname = '';
1401     $atttype = false;
1402     $attary = Array();
1403
1404     while ($pos <= strlen($body)){
1405         $pos = sq_skipspace($body, $pos);
1406         if ($pos == strlen($body)){
1407             /**
1408              * Non-closed tag.
1409              */
1410             return Array(false, false, false, $lt, $pos);
1411         }
1412         /**
1413          * See if we arrived at a ">" or "/>", which means that we reached
1414          * the end of the tag.
1415          */
1416         $matches = Array();
1417         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1418             /**
1419              * Yep. So we did.
1420              */
1421             $pos += strlen($matches{1});
1422             if ($matches{2} == "/>"){
1423                 $tagtype = 3;
1424                 $pos++;
1425             }
1426             return Array($tagname, $attary, $tagtype, $lt, $pos);
1427         }
1428
1429         /**
1430          * There are several types of attributes, with optional
1431          * [:space:] between members.
1432          * Type 1:
1433          *   attrname[:space:]=[:space:]'CDATA'
1434          * Type 2:
1435          *   attrname[:space:]=[:space:]"CDATA"
1436          * Type 3:
1437          *   attr[:space:]=[:space:]CDATA
1438          * Type 4:
1439          *   attrname
1440          *
1441          * We leave types 1 and 2 the same, type 3 we check for
1442          * '"' and convert to "&quot" if needed, then wrap in
1443          * double quotes. Type 4 we convert into:
1444          * attrname="yes".
1445          */
1446         $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1447         if ($regary == false){
1448             /**
1449              * Looks like body ended before the end of tag.
1450              */
1451             return Array(false, false, false, $lt, strlen($body));
1452         }
1453         list($pos, $attname, $match) = $regary;
1454         $attname = strtolower($attname);
1455         /**
1456          * We arrived at the end of attribute name. Several things possible
1457          * here:
1458          * '>'  means the end of the tag and this is attribute type 4
1459          * '/'  if followed by '>' means the same thing as above
1460          * '\s' means a lot of things -- look what it's followed by.
1461          *      anything else means the attribute is invalid.
1462          */
1463         switch($match){
1464             case '/':
1465                 /**
1466                  * This is an xhtml-style tag with a closing / at the
1467                  * end, like so: <img src="blah" />. Check if it's followed
1468                  * by the closing bracket. If not, then this tag is invalid
1469                  */
1470                 if (substr($body, $pos, 2) == "/>"){
1471                     $pos++;
1472                     $tagtype = 3;
1473                 } else {
1474                     $gt = sq_findnxstr($body, $pos, ">");
1475                     $retary = Array(false, false, false, $lt, $gt);
1476                     return $retary;
1477                 }
1478             case '>':
1479                 $attary{$attname} = '"yes"';
1480                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1481                 break;
1482             default:
1483                 /**
1484                  * Skip whitespace and see what we arrive at.
1485                  */
1486                 $pos = sq_skipspace($body, $pos);
1487                 $char = substr($body, $pos, 1);
1488                 /**
1489                  * Two things are valid here:
1490                  * '=' means this is attribute type 1 2 or 3.
1491                  * \w means this was attribute type 4.
1492                  * anything else we ignore and re-loop. End of tag and
1493                  * invalid stuff will be caught by our checks at the beginning
1494                  * of the loop.
1495                  */
1496                 if ($char == "="){
1497                     $pos++;
1498                     $pos = sq_skipspace($body, $pos);
1499                     /**
1500                      * Here are 3 possibilities:
1501                      * "'"  attribute type 1
1502                      * '"'  attribute type 2
1503                      * everything else is the content of tag type 3
1504                      */
1505                     $quot = substr($body, $pos, 1);
1506                     if ($quot == "'"){
1507                         $regary = sq_findnxreg($body, $pos+1, "\'");
1508                         if ($regary == false){
1509                             return Array(false, false, false, $lt, strlen($body));
1510                         }
1511                         list($pos, $attval, $match) = $regary;
1512                         $pos++;
1513                         $attary{$attname} = "'" . $attval . "'";
1514                     } else if ($quot == '"'){
1515                         $regary = sq_findnxreg($body, $pos+1, '\"');
1516                         if ($regary == false){
1517                             return Array(false, false, false, $lt, strlen($body));
1518                         }
1519                         list($pos, $attval, $match) = $regary;
1520                         $pos++;
1521                         $attary{$attname} = '"' . $attval . '"';
1522                     } else {
1523                         /**
1524                          * These are hateful. Look for \s, or >.
1525                          */
1526                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1527                         if ($regary == false){
1528                             return Array(false, false, false, $lt, strlen($body));
1529                         }
1530                         list($pos, $attval, $match) = $regary;
1531                         /**
1532                          * If it's ">" it will be caught at the top.
1533                          */
1534                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1535                         $attary{$attname} = '"' . $attval . '"';
1536                     }
1537                 } else if (preg_match("|[\w/>]|", $char)) {
1538                     /**
1539                      * That was attribute type 4.
1540                      */
1541                     $attary{$attname} = '"yes"';
1542                 } else {
1543                     /**
1544                      * An illegal character. Find next '>' and return.
1545                      */
1546                     $gt = sq_findnxstr($body, $pos, ">");
1547                     return Array(false, false, false, $lt, $gt);
1548                 }
1549                 break;
1550         }
1551     }
1552     /**
1553      * The fact that we got here indicates that the tag end was never
1554      * found. Return invalid tag indication so it gets stripped.
1555      */
1556     return Array(false, false, false, $lt, strlen($body));
1557 }
1558
1559 /**
1560  * Translates entities into literal values so they can be checked.
1561  *
1562  * @param $attvalue the by-ref value to check.
1563  * @param $regex    the regular expression to check against.
1564  * @param $hex      whether the entites are hexadecimal.
1565  * @return          True or False depending on whether there were matches.
1566  */
1567 function sq_deent(&$attvalue, $regex, $hex=false){
1568     $me = 'sq_deent';
1569     $ret_match = false;
1570     preg_match_all($regex, $attvalue, $matches);
1571     if (is_array($matches) && sizeof($matches[0]) > 0){
1572         $repl = Array();
1573         for ($i = 0; $i < sizeof($matches[0]); $i++){
1574             $numval = $matches[1][$i];
1575             if ($hex){
1576                 $numval = hexdec($numval);
1577             }
1578             $repl{$matches[0][$i]} = chr($numval);
1579         }
1580         $attvalue = strtr($attvalue, $repl);
1581         return true;
1582     } else {
1583         return false;
1584     }
1585 }
1586
1587 /**
1588  * This function runs various checks against the attributes.
1589  *
1590  * @param  $tagname         String with the name of the tag.
1591  * @param  $attary          Array with all tag attributes.
1592  * @param  $rm_attnames     See description for sq_sanitize
1593  * @param  $bad_attvals     See description for sq_sanitize
1594  * @param  $add_attr_to_tag See description for sq_sanitize
1595  * @param  $message         message object
1596  * @param  $id              message id
1597  * @return                  Array with modified attributes.
1598  */
1599 function sq_fixatts($tagname,
1600                     $attary,
1601                     $rm_attnames,
1602                     $bad_attvals,
1603                     $add_attr_to_tag,
1604                     $message,
1605                     $id,
1606                     $mailbox
1607                     ){
1608     $me = 'sq_fixatts';
1609     while (list($attname, $attvalue) = each($attary)){
1610         /**
1611          * See if this attribute should be removed.
1612          */
1613         foreach ($rm_attnames as $matchtag=>$matchattrs){
1614             if (preg_match($matchtag, $tagname)){
1615                 foreach ($matchattrs as $matchattr){
1616                     if (preg_match($matchattr, $attname)){
1617                         unset($attary{$attname});
1618                         continue;
1619                     }
1620                 }
1621             }
1622         }
1623         /**
1624          * Remove any backslashes, entities, and extraneous whitespace.
1625          */
1626         sq_defang($attvalue);
1627         sq_unspace($attvalue);
1628
1629         /**
1630          * Now let's run checks on the attvalues.
1631          * I don't expect anyone to comprehend this. If you do,
1632          * get in touch with me so I can drive to where you live and
1633          * shake your hand personally. :)
1634          */
1635         foreach ($bad_attvals as $matchtag=>$matchattrs){
1636             if (preg_match($matchtag, $tagname)){
1637                 foreach ($matchattrs as $matchattr=>$valary){
1638                     if (preg_match($matchattr, $attname)){
1639                         /**
1640                          * There are two arrays in valary.
1641                          * First is matches.
1642                          * Second one is replacements
1643                          */
1644                         list($valmatch, $valrepl) = $valary;
1645                         $newvalue =
1646                             preg_replace($valmatch, $valrepl, $attvalue);
1647                         if ($newvalue != $attvalue){
1648                             $attary{$attname} = $newvalue;
1649                         }
1650                     }
1651                 }
1652             }
1653         }
1654
1655         /**
1656          * Replace empty src tags with the blank image.  src is only used
1657          * for frames, images, and image inputs.  Doing a replace should
1658          * not affect them working as should be, however it will stop
1659          * IE from being kicked off when src for img tags are not set
1660          */
1661         if (($attname == 'src') && ($attvalue == '""')) {
1662             $attary{$attname} = '"' . SM_PATH . 'images/blank.png"';
1663         }
1664
1665         /**
1666          * Turn cid: urls into http-friendly ones.
1667          */
1668         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1669             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1670         }
1671
1672         /**
1673          * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1674          * One day MS might actually make it match something useful, for now, falling
1675          * back to using cid2http, so we can grab the blank.png.
1676          */
1677         if (preg_match("/^[\'\"]\s*outbind:\/\//si", $attvalue)) {
1678             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1679         }
1680
1681     }
1682     /**
1683      * See if we need to append any attributes to this tag.
1684      */
1685     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1686         if (preg_match($matchtag, $tagname)){
1687             $attary = array_merge($attary, $addattary);
1688         }
1689     }
1690     return $attary;
1691 }
1692
1693 /**
1694  * This function edits the style definition to make them friendly and
1695  * usable in SquirrelMail.
1696  *
1697  * @param  $message  the message object
1698  * @param  $id       the message id
1699  * @param  $content  a string with whatever is between <style> and </style>
1700  * @param  $mailbox  the message mailbox
1701  * @return           a string with edited content.
1702  */
1703 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1704     global $view_unsafe_images;
1705     $me = 'sq_fixstyle';
1706     $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1707     if ($ret == FALSE){
1708         return array(FALSE, strlen($body));
1709     }
1710     $newpos = $ret[0] + strlen($ret[2]);
1711     $content = $ret[1];
1712     /**
1713     * First look for general BODY style declaration, which would be
1714     * like so:
1715     * body {background: blah-blah}
1716     * and change it to .bodyclass so we can just assign it to a <div>
1717     */
1718     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1719     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1720     /**
1721     * Fix url('blah') declarations.
1722     */
1723     //   $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1724     //                           "url(\\1$secremoveimg\\2)", $content);
1725     // remove NUL
1726     $content = str_replace("\0", "", $content);
1727     // translate ur\l and variations (IE parses that)
1728     $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
1729     // NB I insert NUL characters to keep to avoid an infinite loop. They are removed after the loop.
1730     while (preg_match("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si", $content, $matches)) {
1731         $sProto = strtolower($matches[1]);
1732         switch ($sProto) {
1733             /**
1734              * Fix url('https*://.*) declarations but only if $view_unsafe_images
1735              * is false.
1736              */
1737             case 'https':
1738             case 'http':
1739                 if (!$view_unsafe_images){
1740
1741                     $sExpr = "/url\s*\(\s*[\'\"]?\s*$sProto*:.*[\'\"]?\s*\)/si";
1742                     $content = preg_replace($sExpr, "u\0r\0l(\\1$secremoveimg\\2)", $content);
1743
1744                 } else {
1745                     $content = preg_replace('/url/i',"u\0r\0l",$content);
1746                 }
1747                 break;
1748             /**
1749              * Fix urls that refer to cid:
1750              */
1751             case 'cid':
1752                 $cidurl = 'cid:'. $matches[2];
1753                 $httpurl = sq_cid2http($message, $id, $cidurl, $mailbox);
1754                 // escape parentheses that can modify the regular expression
1755                 $cidurl = str_replace(array('(',')'),array('\\(','\\)'),$cidurl);
1756                 $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1757                                         "u\0r\0l($httpurl)", $content);
1758                 break;
1759             default:
1760                 /**
1761                  * replace url with protocol other then the white list
1762                  * http,https and cid by an empty string.
1763                  */
1764                 $content = preg_replace("/url\s*\(\s*[\'\"]?([^:]+):(.*)?[\'\"]?\s*\)/si",
1765                                  "", $content);
1766                 break;
1767         }
1768     }
1769     // remove NUL
1770     $content = str_replace("\0", "", $content);
1771    /**
1772     * Remove any backslashes, entities, and extraneous whitespace.
1773     */
1774     $contentTemp = $content;
1775     sq_defang($contentTemp);
1776     sq_unspace($contentTemp);
1777
1778     /**
1779      * Fix stupid css declarations which lead to vulnerabilities
1780      * in IE.
1781      */
1782     $match   = Array('/\/\*.*\*\//',
1783                     '/expression/i',
1784                     '/behaviou*r/i',
1785                     '/binding/i',
1786                     '/include-source/i');
1787     $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy');
1788     $contentNew = preg_replace($match, $replace, $contentTemp);
1789     if ($contentNew !== $contentTemp) {
1790         // insecure css declarations are used. From now on we don't care
1791         // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
1792         $content = $contentNew;
1793     }
1794     return array($content, $newpos);
1795 }
1796
1797
1798 /**
1799  * This function converts cid: url's into the ones that can be viewed in
1800  * the browser.
1801  *
1802  * @param  $message  the message object
1803  * @param  $id       the message id
1804  * @param  $cidurl   the cid: url.
1805  * @param  $mailbox  the message mailbox
1806  * @return           a string with a http-friendly url
1807  */
1808 function sq_cid2http($message, $id, $cidurl, $mailbox){
1809     /**
1810      * Get rid of quotes.
1811      */
1812     $quotchar = substr($cidurl, 0, 1);
1813     if ($quotchar == '"' || $quotchar == "'"){
1814         $cidurl = str_replace($quotchar, "", $cidurl);
1815     } else {
1816         $quotchar = '';
1817     }
1818     $cidurl = substr(trim($cidurl), 4);
1819
1820     $match_str = '/\{.*?\}\//';
1821     $str_rep = '';
1822     $cidurl = preg_replace($match_str, $str_rep, $cidurl);
1823
1824     $linkurl = find_ent_id($cidurl, $message);
1825     /* in case of non-save cid links $httpurl should be replaced by a sort of
1826        unsave link image */
1827     $httpurl = '';
1828
1829     /**
1830      * This is part of a fix for Outlook Express 6.x generating
1831      * cid URLs without creating content-id headers. These images are
1832      * not part of the multipart/related html mail. The html contains
1833      * <img src="cid:{some_id}/image_filename.ext"> references to
1834      * attached images with as goal to render them inline although
1835      * the attachment disposition property is not inline.
1836      */
1837
1838     if (empty($linkurl)) {
1839         if (preg_match('/{.*}\//', $cidurl)) {
1840             $cidurl = preg_replace('/{.*}\//','', $cidurl);
1841             if (!empty($cidurl)) {
1842                 $linkurl = find_ent_id($cidurl, $message);
1843             }
1844         }
1845     }
1846
1847     if (!empty($linkurl)) {
1848         $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1849             "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1850             '&amp;ent_id=' . $linkurl . $quotchar;
1851     } else {
1852         /**
1853          * If we couldn't generate a proper img url, drop in a blank image
1854          * instead of sending back empty, otherwise it causes unusual behaviour
1855          */
1856         $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
1857     }
1858
1859     return $httpurl;
1860 }
1861
1862 /**
1863  * This function changes the <body> tag into a <div> tag since we
1864  * can't really have a body-within-body.
1865  *
1866  * @param  $attary   an array of attributes and values of <body>
1867  * @param  $mailbox  mailbox we're currently reading (for cid2http)
1868  * @param  $message  current message (for cid2http)
1869  * @param  $id       current message id (for cid2http)
1870  * @return           a modified array of attributes to be set for <div>
1871  */
1872 function sq_body2div($attary, $mailbox, $message, $id){
1873     $me = 'sq_body2div';
1874     $divattary = Array('class' => "'bodyclass'");
1875     $text = '#000000';
1876     $has_bgc_stl = $has_txt_stl = false;
1877     $styledef = '';
1878     if (is_array($attary) && sizeof($attary) > 0){
1879         foreach ($attary as $attname=>$attvalue){
1880             $quotchar = substr($attvalue, 0, 1);
1881             $attvalue = str_replace($quotchar, "", $attvalue);
1882             switch ($attname){
1883                 case 'background':
1884                     $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1885                     $styledef .= "background-image: url('$attvalue'); ";
1886                     break;
1887                 case 'bgcolor':
1888                     $has_bgc_stl = true;
1889                     $styledef .= "background-color: $attvalue; ";
1890                     break;
1891                 case 'text':
1892                     $has_txt_stl = true;
1893                     $styledef .= "color: $attvalue; ";
1894                     break;
1895             }
1896         }
1897         // Outlook defines a white bgcolor and no text color. This can lead to
1898         // white text on a white bg with certain themes.
1899         if ($has_bgc_stl && !$has_txt_stl) {
1900             $styledef .= "color: $text; ";
1901         }
1902         if (strlen($styledef) > 0){
1903             $divattary{"style"} = "\"$styledef\"";
1904         }
1905     }
1906     return $divattary;
1907 }
1908
1909 /**
1910  * This is the main function and the one you should actually be calling.
1911  * There are several variables you should be aware of an which need
1912  * special description.
1913  *
1914  * Since the description is quite lengthy, see it here:
1915  * http://linux.duke.edu/projects/mini/htmlfilter/
1916  *
1917  * @param $body                 the string with HTML you wish to filter
1918  * @param $tag_list             see description above
1919  * @param $rm_tags_with_content see description above
1920  * @param $self_closing_tags    see description above
1921  * @param $force_tag_closing    see description above
1922  * @param $rm_attnames          see description above
1923  * @param $bad_attvals          see description above
1924  * @param $add_attr_to_tag      see description above
1925  * @param $message              message object
1926  * @param $id                   message id
1927  * @return                      sanitized html safe to show on your pages.
1928  */
1929 function sq_sanitize($body,
1930                      $tag_list,
1931                      $rm_tags_with_content,
1932                      $self_closing_tags,
1933                      $force_tag_closing,
1934                      $rm_attnames,
1935                      $bad_attvals,
1936                      $add_attr_to_tag,
1937                      $message,
1938                      $id,
1939                      $mailbox
1940                      ){
1941     $me = 'sq_sanitize';
1942     $rm_tags = array_shift($tag_list);
1943     /**
1944      * Normalize rm_tags and rm_tags_with_content.
1945      */
1946     @array_walk($tag_list, 'sq_casenormalize');
1947     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1948     @array_walk($self_closing_tags, 'sq_casenormalize');
1949     /**
1950      * See if tag_list is of tags to remove or tags to allow.
1951      * false  means remove these tags
1952      * true   means allow these tags
1953      */
1954     $curpos = 0;
1955     $open_tags = Array();
1956     $trusted = "\n<!-- begin sanitized html -->\n";
1957     $skip_content = false;
1958     /**
1959      * Take care of netscape's stupid javascript entities like
1960      * &{alert('boo')};
1961      */
1962     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1963
1964     while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1965         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1966         $free_content = substr($body, $curpos, $lt-$curpos);
1967         /**
1968          * Take care of <style>
1969          */
1970         if ($tagname == "style" && $tagtype == 1){
1971             list($free_content, $curpos) =
1972                 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
1973             if ($free_content != FALSE){
1974                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1975                 $trusted .= $free_content;
1976                 $trusted .= sq_tagprint($tagname, false, 2);
1977             }
1978             continue;
1979         }
1980         if ($skip_content == false){
1981             $trusted .= $free_content;
1982         }
1983         if ($tagname != FALSE){
1984             if ($tagtype == 2){
1985                 if ($skip_content == $tagname){
1986                     /**
1987                      * Got to the end of tag we needed to remove.
1988                      */
1989                     $tagname = false;
1990                     $skip_content = false;
1991                 } else {
1992                     if ($skip_content == false){
1993                         if ($tagname == "body"){
1994                             $tagname = "div";
1995                         }
1996                         if (isset($open_tags{$tagname}) &&
1997                                 $open_tags{$tagname} > 0){
1998                             $open_tags{$tagname}--;
1999                         } else {
2000                             $tagname = false;
2001                         }
2002                     }
2003                 }
2004             } else {
2005                 /**
2006                  * $rm_tags_with_content
2007                  */
2008                 if ($skip_content == false){
2009                     /**
2010                      * See if this is a self-closing type and change
2011                      * tagtype appropriately.
2012                      */
2013                     if ($tagtype == 1
2014                             && in_array($tagname, $self_closing_tags)){
2015                         $tagtype = 3;
2016                     }
2017                     /**
2018                      * See if we should skip this tag and any content
2019                      * inside it.
2020                      */
2021                     if ($tagtype == 1 &&
2022                             in_array($tagname, $rm_tags_with_content)){
2023                         $skip_content = $tagname;
2024                     } else {
2025                         if (($rm_tags == false
2026                                     && in_array($tagname, $tag_list)) ||
2027                                 ($rm_tags == true &&
2028                                  !in_array($tagname, $tag_list))){
2029                             $tagname = false;
2030                         } else {
2031                             /**
2032                              * Convert body into div.
2033                              */
2034                             if ($tagname == "body"){
2035                                 $tagname = "div";
2036                                 $attary = sq_body2div($attary, $mailbox,
2037                                         $message, $id);
2038                             }
2039                             if ($tagtype == 1){
2040                                 if (isset($open_tags{$tagname})){
2041                                     $open_tags{$tagname}++;
2042                                 } else {
2043                                     $open_tags{$tagname}=1;
2044                                 }
2045                             }
2046                             /**
2047                              * This is where we run other checks.
2048                              */
2049                             if (is_array($attary) && sizeof($attary) > 0){
2050                                 $attary = sq_fixatts($tagname,
2051                                                      $attary,
2052                                                      $rm_attnames,
2053                                                      $bad_attvals,
2054                                                      $add_attr_to_tag,
2055                                                      $message,
2056                                                      $id,
2057                                                      $mailbox
2058                                                      );
2059                             }
2060                         }
2061                     }
2062                 }
2063             }
2064             if ($tagname != false && $skip_content == false){
2065                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2066             }
2067         }
2068         $curpos = $gt+1;
2069     }
2070     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2071     if ($force_tag_closing == true){
2072         foreach ($open_tags as $tagname=>$opentimes){
2073             while ($opentimes > 0){
2074                 $trusted .= '</' . $tagname . '>';
2075                 $opentimes--;
2076             }
2077         }
2078         $trusted .= "\n";
2079     }
2080     $trusted .= "<!-- end sanitized html -->\n";
2081     return $trusted;
2082 }
2083
2084 /**
2085  * This is a wrapper function to call html sanitizing routines.
2086  *
2087  * @param  $body  the body of the message
2088  * @param  $id    the id of the message
2089
2090  * @param  $message
2091  * @param  $mailbox
2092  * @param  boolean $take_mailto_links When TRUE, converts mailto: links
2093  *                                    into internal SM compose links
2094  *                                    (optional; default = TRUE)
2095  * @return        a string with html safe to display in the browser.
2096  */
2097 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2098
2099     // require_once(SM_PATH . 'functions/url_parser.php');  // for $MailTo_PReg_Match
2100
2101     global $attachment_common_show_images, $view_unsafe_images,
2102            $has_unsafe_images;
2103     /**
2104      * Don't display attached images in HTML mode.
2105      */
2106     $attachment_common_show_images = false;
2107     $tag_list = Array(
2108             false,
2109             "object",
2110             "meta",
2111             "html",
2112             "head",
2113             "base",
2114             "link",
2115             "frame",
2116             "iframe",
2117             "plaintext",
2118             "marquee"
2119             );
2120
2121     $rm_tags_with_content = Array(
2122             "script",
2123             "applet",
2124             "embed",
2125             "title",
2126             "frameset",
2127             "xmp",
2128             "xml"
2129             );
2130
2131     $self_closing_tags =  Array(
2132             "img",
2133             "br",
2134             "hr",
2135             "input",
2136             "outbind"
2137             );
2138
2139     $force_tag_closing = true;
2140
2141     $rm_attnames = Array(
2142             "/.*/" =>
2143             Array(
2144                 "/target/i",
2145                 "/^on.*/i",
2146                 "/^dynsrc/i",
2147                 "/^data.*/i",
2148                 "/^lowsrc.*/i"
2149                 )
2150             );
2151
2152     $secremoveimg = "../images/" . _("sec_remove_eng.png");
2153     $bad_attvals = Array(
2154             "/.*/" =>
2155             Array(
2156                 "/^src|background/i" =>
2157                 Array(
2158                     Array(
2159                         "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2160                         "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2161                         "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2162                         ),
2163                     Array(
2164                         "\\1$secremoveimg\\2",
2165                         "\\1$secremoveimg\\2",
2166                         "\\1$secremoveimg\\2",
2167                         )
2168                     ),
2169                 "/^href|action/i" =>
2170                 Array(
2171                     Array(
2172                         "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2173                         "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2174                         "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2175                         ),
2176                     Array(
2177                         "\\1#\\1",
2178                         "\\1#\\1",
2179                         "\\1#\\1"
2180                         )
2181                     ),
2182         "/^style/i" =>
2183             Array(
2184                 Array(
2185                     "/\/\*.*\*\//",
2186                     "/expression/i",
2187                     "/binding/i",
2188                     "/behaviou*r/i",
2189                     "/include-source/i",
2190                     "/position\s*:\s*absolute/i",
2191                     "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
2192                     "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2193                     "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2194                     "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2195                     "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2196                     ),
2197                 Array(
2198                     "",
2199                     "idiocy",
2200                     "idiocy",
2201                     "idiocy",
2202                     "idiocy",
2203                     "",
2204                     "url",
2205                     "url(\\1#\\1)",
2206                     "url(\\1#\\1)",
2207                     "url(\\1#\\1)",
2208                     "\\1:url(\\2#\\3)"
2209                     )
2210                 )
2211             )
2212         );
2213     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
2214         $view_unsafe_images = false;
2215     }
2216     if (!$view_unsafe_images){
2217         /**
2218          * Remove any references to http/https if view_unsafe_images set
2219          * to false.
2220          */
2221         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2222                 '/^([\'\"])\s*https*:.*([\'\"])/si');
2223         array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2224                 "\\1$secremoveimg\\1");
2225         array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2226                 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2227         array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2228                 "url(\\1$secremoveimg\\1)");
2229     }
2230
2231     $add_attr_to_tag = Array(
2232             "/^a$/i" =>
2233             Array('target'=>'"_blank"',
2234                 'title'=>'"'._("This external link will open in a new window").'"'
2235                 )
2236             );
2237     $trusted = sq_sanitize($body,
2238                            $tag_list,
2239                            $rm_tags_with_content,
2240                            $self_closing_tags,
2241                            $force_tag_closing,
2242                            $rm_attnames,
2243                            $bad_attvals,
2244                            $add_attr_to_tag,
2245                            $message,
2246                            $id,
2247                            $mailbox
2248                            );
2249     if (preg_match("|$secremoveimg|i", $trusted)){
2250         $has_unsafe_images = true;
2251     }
2252
2253     // we want to parse mailto's in HTML output, change to SM compose links
2254     // this is a modified version of code from url_parser.php... but Marc is
2255     // right: we need a better filtering implementation; adding this randomly
2256     // here is not a great solution
2257     //
2258     if ($take_mailto_links) {
2259         // parseUrl($trusted);   // this even parses URLs inside of tags... too aggressive
2260         global $MailTo_PReg_Match;
2261         $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1) ;
2262         if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2263             foreach ($regs[0] as $i => $mailto_before) {
2264                 $mailto_params = $regs[10][$i];
2265                 // get rid of any tailing quote since we have to add send_to to the end
2266                 //
2267                 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2268                     $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2269                 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2270                     $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2271
2272                 if ($regs[1][$i]) {    //if there is an email addr before '?', we need to merge it with the params
2273                     $to = 'to=' . $regs[1][$i];
2274                     if (strpos($mailto_params, 'to=') > -1)    //already a 'to='
2275                         $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2276                     else {
2277                         if ($mailto_params)    //already some params, append to them
2278                             $mailto_params .= '&amp;' . $to;
2279                         else
2280                             $mailto_params .= '?' . $to;
2281                     }
2282                 }
2283
2284                 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2285
2286                 // we'll already have target=_blank, no need to allow comp_in_new
2287                 // here (which would be a lot more work anyway)
2288                 //
2289                 global $compose_new_win;
2290                 $temp_comp_in_new = $compose_new_win;
2291                 $compose_new_win = 0;
2292                 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2293                 $compose_new_win = $temp_comp_in_new;
2294
2295                 // remove <a href=" and anything after the next quote (we only
2296                 // need the uri, not the link HTML) in compose uri
2297                 //
2298                 $comp_uri = substr($comp_uri, 9);
2299                 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2300                 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2301             }
2302         }
2303     }
2304
2305     return $trusted;
2306 }
2307
2308 /**
2309  * function SendDownloadHeaders - send file to the browser
2310  *
2311  * Original Source: SM core src/download.php
2312  * moved here to make it available to other code, and separate
2313  * front end from back end functionality.
2314  *
2315  * @param string $type0 first half of mime type
2316  * @param string $type1 second half of mime type
2317  * @param string $filename filename to tell the browser for downloaded file
2318  * @param boolean $force whether to force the download dialog to pop
2319  * @param optional integer $filesize send the Content-Header and length to the browser
2320  * @return void
2321  */
2322 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2323     global $languages, $squirrelmail_language;
2324     $isIE = $isIE6plus = false;
2325
2326     sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2327
2328     if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2329             strstr($HTTP_USER_AGENT, 'Opera') === false) {
2330         $isIE = true;
2331     }
2332
2333     if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2334         ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2335         $isIE6plus = true;
2336     }
2337
2338     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2339             function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2340         $filename =
2341             call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2342     } else {
2343         $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
2344     }
2345
2346     // A Pox on Microsoft and it's Internet Explorer!
2347     //
2348     // IE has lots of bugs with file downloads.
2349     // It also has problems with SSL.  Both of these cause problems
2350     // for us in this function.
2351     //
2352     // See this article on Cache Control headers and SSL
2353     // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2354     //
2355     // The best thing you can do for IE is to upgrade to the latest
2356     // version
2357     //set all the Cache Control Headers for IE
2358     if ($isIE) {
2359         $filename=rawurlencode($filename);
2360         header ("Pragma: public");
2361         header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2362         header ("Cache-Control: post-check=0, pre-check=0", false);
2363         header ("Cache-Control: private");
2364
2365         //set the inline header for IE, we'll add the attachment header later if we need it
2366         header ("Content-Disposition: inline; filename=$filename");
2367     }
2368
2369     if (!$force) {
2370         // Try to show in browser window
2371         header ("Content-Disposition: inline; filename=\"$filename\"");
2372         header ("Content-Type: $type0/$type1; name=\"$filename\"");
2373     } else {
2374         // Try to pop up the "save as" box
2375
2376         // IE makes this hard.  It pops up 2 save boxes, or none.
2377         // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2378         // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2379         // But, according to Microsoft, it is "RFC compliant but doesn't
2380         // take into account some deviations that allowed within the
2381         // specification."  Doesn't that mean RFC non-compliant?
2382         // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2383
2384         // all browsers need the application/octet-stream header for this
2385         header ("Content-Type: application/octet-stream; name=\"$filename\"");
2386
2387         // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2388         // Do not have quotes around filename, but that applied to
2389         // "attachment"... does it apply to inline too?
2390         header ("Content-Disposition: attachment; filename=\"$filename\"");
2391
2392         if ($isIE && !$isIE6plus) {
2393             // This combination seems to work mostly.  IE 5.5 SP 1 has
2394             // known issues (see the Microsoft Knowledge Base)
2395
2396             // This works for most types, but doesn't work with Word files
2397             header ("Content-Type: application/download; name=\"$filename\"");
2398
2399             // These are spares, just in case.  :-)
2400             //header("Content-Type: $type0/$type1; name=\"$filename\"");
2401             //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2402             //header("Content-Type: application/octet-stream; name=\"$filename\"");
2403         } else {
2404             // another application/octet-stream forces download for Netscape
2405             header ("Content-Type: application/octet-stream; name=\"$filename\"");
2406         }
2407     }
2408
2409     //send the content-length header if the calling function provides it
2410     if ($filesize > 0) {
2411         header("Content-Length: $filesize");
2412     }
2413
2414 }  // end fn SendDownloadHeaders