functions/mime.php

   1 <?php
   2
   3 /**
   4  * mime.php
   5  *
   6  * Copyright (c) 1999-2003 The SquirrelMail Project Team
   7  * Licensed under the GNU GPL. For full terms see the file COPYING.
   8  *
   9  * This contains the functions necessary to detect and decode MIME
  10  * messages.
  11  *
  12  * $Id$
  13  * @package squirrelmail
  14  */
  15
  16 /** The typical includes... */
  17 require_once(SM_PATH . 'functions/imap.php');
  18 require_once(SM_PATH . 'functions/attachment_common.php');
  19
  20 /* -------------------------------------------------------------------------- */
  21 /* MIME DECODING                                                              */
  22 /* -------------------------------------------------------------------------- */
  23
  24 /**
  25  * Get the MIME structure
  26  *
  27  * This function gets the structure of a message and stores it in the "message" class.
  28  * It will return this object for use with all relevant header information and
  29  * fully parsed into the standard "message" object format.
  30  */
  31 function mime_structure ($bodystructure, $flags=array()) {
  32
  33     /* Isolate the body structure and remove beginning and end parenthesis. */
  34     $read = trim(substr ($bodystructure, strpos(strtolower($bodystructure), 'bodystructure') + 13));
  35     $read = trim(substr ($read, 0, -1));
  36     $i = 0;
  37     $msg = Message::parseStructure($read,$i);
  38     if (!is_object($msg)) {
  39         include_once(SM_PATH . 'functions/display_messages.php');
  40         global $color, $mailbox;
  41         /* removed urldecode because $_GET is auto urldecoded ??? */
  42         displayPageHeader( $color, $mailbox );
  43         $errormessage  = _("SquirrelMail could not decode the bodystructure of the message");
  44         $errormessage .= '<BR>'._("the provided bodystructure by your imap-server").':<BR><BR>';
  45         $errormessage .= '<pre>' . htmlspecialchars($read) . '</pre>';
  46         plain_error_message( $errormessage, $color );
  47         echo '</body></html>';
  48         exit;
  49     }
  50     if (count($flags)) {
  51         foreach ($flags as $flag) {
  52             $char = strtoupper($flag{1});
  53             switch ($char) {
  54                 case 'S':
  55                     if (strtolower($flag) == '\\seen') {
  56                         $msg->is_seen = true;
  57                     }
  58                     break;
  59                 case 'A':
  60                     if (strtolower($flag) == '\\answered') {
  61                         $msg->is_answered = true;
  62                     }
  63                     break;
  64                 case 'D':
  65                     if (strtolower($flag) == '\\deleted') {
  66                         $msg->is_deleted = true;
  67                     }
  68                     break;
  69                 case 'F':
  70                     if (strtolower($flag) == '\\flagged') {
  71                         $msg->is_flagged = true;
  72                     }
  73                     break;
  74                 case 'M':
  75                     if (strtolower($flag) == '$mdnsent') {
  76                         $msg->is_mdnsent = true;
  77                     }
  78                     break;
  79                 default:
  80                     break;
  81             }
  82         }
  83     }
  84     //    listEntities($msg);
  85     return $msg;
  86 }
  87
  88
  89
  90 /* This starts the parsing of a particular structure.  It is called recursively,
  91  * so it can be passed different structures.  It returns an object of type
  92  * $message.
  93  * First, it checks to see if it is a multipart message.  If it is, then it
  94  * handles that as it sees is necessary.  If it is just a regular entity,
  95  * then it parses it and adds the necessary header information (by calling out
  96  * to mime_get_elements()
  97  */
  98
  99 function mime_fetch_body($imap_stream, $id, $ent_id=1, $fetch_size=0) {
 100     global $uid_support;
 101     /* Do a bit of error correction.  If we couldn't find the entity id, just guess
 102      * that it is the first one.  That is usually the case anyway.
 103      */
 104
 105     if (!$ent_id) {
 106         $cmd = "FETCH $id BODY[]";
 107     } else {
 108         $cmd = "FETCH $id BODY[$ent_id]";
 109     }
 110
 111     if ($fetch_size!=0) $cmd .= "<0.$fetch_size>";
 112
 113     $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, $uid_support);
 114     do {
 115         $topline = trim(array_shift($data));
 116     } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
 117
 118     $wholemessage = implode('', $data);
 119     if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
 120         $ret = substr($wholemessage, 0, $regs[1]);
 121         /* There is some information in the content info header that could be important
 122          * in order to parse html messages. Let's get them here.
 123          */
 124 //        if ($ret{0} == '<') {
 125 //            $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, $uid_support);
 126 //        }
 127     } else if (ereg('"([^"]*)"', $topline, $regs)) {
 128         $ret = $regs[1];
 129     } else {
 130         global $where, $what, $mailbox, $passed_id, $startMessage;
 131         $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
 132         if (isset($where) && isset($what)) {
 133             $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
 134         } else {
 135             $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 136         }
 137         $par .= '&amp;response=' . urlencode($response) .
 138                 '&amp;message='  . urlencode($message)  .
 139                 '&amp;topline='  . urlencode($topline);
 140
 141         echo   '<tt><br>' .
 142                '<table width="80%"><tr>' .
 143                '<tr><td colspan=2>' .
 144                _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
 145                '</td></tr>' .
 146                '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
 147                '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
 148                '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
 149                '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
 150                "</table><BR></tt></font><hr>";
 151
 152         $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, $uid_support);
 153         array_shift($data);
 154         $wholemessage = implode('', $data);
 155
 156         $ret = $wholemessage;
 157     }
 158     return $ret;
 159 }
 160
 161 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding) {
 162     global $uid_support;
 163
 164     /* Don't kill the connection if the browser is over a dialup
 165      * and it would take over 30 seconds to download it.
 166      * Dont call set_time_limit in safe mode.
 167      */
 168
 169     if (!ini_get('safe_mode')) {
 170         set_time_limit(0);
 171     }
 172     /* in case of base64 encoded attachments, do not buffer them.
 173        Instead, echo the decoded attachment directly to screen */
 174     if (strtolower($encoding) == 'base64') {
 175         if (!$ent_id) {
 176            $query = "FETCH $id BODY[]";
 177         } else {
 178            $query = "FETCH $id BODY[$ent_id]";
 179         }
 180         sqimap_run_command($imap_stream,$query,true,$response,$message,$uid_support,'sqimap_base64_decode','php://stdout',true);
 181     } else {
 182        $body = mime_fetch_body ($imap_stream, $id, $ent_id);
 183        echo decodeBody($body, $encoding);
 184     }
 185
 186     /*
 187        TODO, use the same method for quoted printable.
 188        However, I assume that quoted printable attachments aren't that large
 189        so the performancegain / memory usage drop will be minimal.
 190        If we decide to add that then we need to adapt sqimap_fread because
 191        we need to split te result on \n and fread doesn't stop at \n. That
 192        means we also should provide $results from sqimap_fread (by ref) to
 193        te function and set $no_return to false. The $filter function for
 194        quoted printable should handle unsetting of $results.
 195     */
 196     /*
 197        TODO 2: find out how we write to the output stream php://stdout. fwrite
 198        doesn't work because 'php://stdout isn't a stream.
 199     */
 200
 201     return;
 202 /*
 203     fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
 204     $cnt = 0;
 205     $continue = true;
 206     $read = fgets ($imap_stream,8192);
 207
 208
 209     // This could be bad -- if the section has sqimap_session_id() . ' OK'
 210     // or similar, it will kill the download.
 211     while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
 212         if (trim($read) == ')==') {
 213             $read1 = $read;
 214             $read = fgets ($imap_stream,4096);
 215             if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
 216                 return;
 217             } else {
 218                 echo decodeBody($read1, $encoding) .
 219                      decodeBody($read, $encoding);
 220             }
 221         } else if ($cnt) {
 222             echo decodeBody($read, $encoding);
 223         }
 224         $read = fgets ($imap_stream,4096);
 225         $cnt++;
 226 //      break;
 227     }
 228 */
 229 }
 230
 231 /* -[ END MIME DECODING ]----------------------------------------------------------- */
 232
 233 /* This is here for debugging purposes.  It will print out a list
 234  * of all the entity IDs that are in the $message object.
 235  */
 236 function listEntities ($message) {
 237     if ($message) {
 238         echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br>';
 239         for ($i = 0; isset($message->entities[$i]); $i++) {
 240             echo "$i : ";
 241             $msg = listEntities($message->entities[$i]);
 242
 243             if ($msg) {
 244                 echo "return: ";
 245                 return $msg;
 246             }
 247         }
 248     }
 249 }
 250
 251 function getPriorityStr($priority) {
 252     $priority_level = substr($priority,0,1);
 253
 254     switch($priority_level) {
 255         /* Check for a higher then normal priority. */
 256         case '1':
 257         case '2':
 258             $priority_string = _("High");
 259             break;
 260
 261         /* Check for a lower then normal priority. */
 262         case '4':
 263         case '5':
 264             $priority_string = _("Low");
 265             break;
 266
 267         /* Check for a normal priority. */
 268         case '3':
 269         default:
 270             $priority_level = '3';
 271             $priority_string = _("Normal");
 272             break;
 273
 274     }
 275     return $priority_string;
 276 }
 277
 278 /* returns a $message object for a particular entity id */
 279 function getEntity ($message, $ent_id) {
 280     return $message->getEntity($ent_id);
 281 }
 282
 283 /* translateText
 284  * Extracted from strings.php 23/03/2002
 285  */
 286
 287 function translateText(&$body, $wrap_at, $charset) {
 288     global $where, $what;   /* from searching */
 289     global $color;          /* color theme */
 290
 291     require_once(SM_PATH . 'functions/url_parser.php');
 292
 293     $body_ary = explode("\n", $body);
 294     for ($i=0; $i < count($body_ary); $i++) {
 295         $line = $body_ary[$i];
 296         if (strlen($line) - 2 >= $wrap_at) {
 297             sqWordWrap($line, $wrap_at);
 298         }
 299         $line = charset_decode($charset, $line);
 300         $line = str_replace("\t", '        ', $line);
 301
 302         parseUrl ($line);
 303
 304         $quotes = 0;
 305         $pos = 0;
 306         $j = strlen($line);
 307
 308         while ($pos < $j) {
 309             if ($line[$pos] == ' ') {
 310                 $pos++;
 311             } else if (strpos($line, '&gt;', $pos) === $pos) {
 312                 $pos += 4;
 313                 $quotes++;
 314             } else {
 315                 break;
 316             }
 317         }
 318
 319         if ($quotes % 2) {
 320             if (!isset($color[13])) {
 321                 $color[13] = '#800000';
 322             }
 323             $line = '<font color="' . $color[13] . '">' . $line . '</font>';
 324         } elseif ($quotes) {
 325             if (!isset($color[14])) {
 326                 $color[14] = '#FF0000';
 327             }
 328             $line = '<font color="' . $color[14] . '">' . $line . '</font>';
 329         }
 330
 331         $body_ary[$i] = $line;
 332     }
 333     $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
 334 }
 335
 336 /* This returns a parsed string called $body. That string can then
 337  * be displayed as the actual message in the HTML. It contains
 338  * everything needed, including HTML Tags, Attachments at the
 339  * bottom, etc.
 340  */
 341 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
 342     /* This if statement checks for the entity to show as the
 343      * primary message. To add more of them, just put them in the
 344      * order that is their priority.
 345      */
 346     global $startMessage, $username, $key, $imapServerAddress, $imapPort,
 347            $show_html_default, $sort, $has_unsafe_images, $passed_ent_id;
 348     global $languages, $squirrelmail_language;
 349
 350     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
 351         $view_unsafe_images = false;
 352     }
 353
 354     $body = '';
 355     $urlmailbox = urlencode($mailbox);
 356     $body_message = getEntity($message, $ent_num);
 357     if (($body_message->header->type0 == 'text') ||
 358         ($body_message->header->type0 == 'rfc822')) {
 359         $body = mime_fetch_body ($imap_stream, $id, $ent_num);
 360         $body = decodeBody($body, $body_message->header->encoding);
 361
 362         if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 363             function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 364             if (mb_detect_encoding($body) != 'ASCII') {
 365                 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
 366             }
 367         }
 368         $hookResults = do_hook("message_body", $body);
 369         $body = $hookResults[1];
 370
 371         /* If there are other types that shouldn't be formatted, add
 372          * them here.
 373          */
 374
 375         if ($body_message->header->type1 == 'html') {
 376             if ($show_html_default <> 1) {
 377                 $entity_conv = array('&nbsp;' => ' ',
 378                                      '<p>'    => "\n",
 379                                      '<br>'   => "\n",
 380                                      '<P>'    => "\n",
 381                                      '<BR>'   => "\n",
 382                                      '&gt;'   => '>',
 383                                      '&lt;'   => '<');
 384                 $body = strtr($body, $entity_conv);
 385                 $body = strip_tags($body);
 386                 $body = trim($body);
 387                 translateText($body, $wrap_at,
 388                               $body_message->header->getParameter('charset'));
 389             } else {
 390                 $body = magicHTML($body, $id, $message, $mailbox);
 391             }
 392         } else {
 393             translateText($body, $wrap_at,
 394                           $body_message->header->getParameter('charset'));
 395         }
 396         $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
 397                 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
 398                 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 399         if (isset($passed_ent_id)) {
 400             $link .= '&amp;passed_ent_id='.$passed_ent_id;
 401         }
 402         $body .= '<center><small><a href="download.php?absolute_dl=true&amp;' .
 403                  $link . '">' . _("Download this as a file") .  '</a>';
 404         if ($view_unsafe_images) {
 405             $text = _("Hide Unsafe Images");
 406         } else {
 407             if (isset($has_unsafe_images) && $has_unsafe_images) {
 408                 $link .= '&amp;view_unsafe_images=1';
 409                 $text = _("View Unsafe Images");
 410             } else {
 411                 $text = '';
 412             }
 413         }
 414         if($text != '') {
 415             $body .= '&nbsp;|&nbsp;<a href="read_body.php?' . $link . '">' . $text . '</a>';
 416         }
 417         $body .= '</small></center><br>' . "\n";
 418     }
 419     return $body;
 420 }
 421
 422
 423 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 424     global $where, $what, $startMessage, $color;
 425     static $ShownHTML = 0;
 426
 427     $att_ar = $message->getAttachments($exclude_id);
 428
 429     if (!count($att_ar)) return '';
 430
 431     $attachments = '';
 432
 433     $urlMailbox = urlencode($mailbox);
 434
 435     foreach ($att_ar as $att) {
 436         $ent = $att->entity_id;
 437         $header = $att->header;
 438         $type0 = strtolower($header->type0);
 439         $type1 = strtolower($header->type1);
 440         $name = '';
 441         $links['download link']['text'] = _("download");
 442         $links['download link']['href'] = SM_PATH .
 443                 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 444         $ImageURL = '';
 445         if ($type0 =='message' && $type1 == 'rfc822') {
 446             $default_page = SM_PATH . 'src/read_body.php';
 447             $rfc822_header = $att->rfc822_header;
 448             $filename = $rfc822_header->subject;
 449             if (trim( $filename ) == '') {
 450                 $filename = 'untitled-[' . $ent . ']' ;
 451             }
 452             $from_o = $rfc822_header->from;
 453             if (is_object($from_o)) {
 454                 $from_name = decodeHeader($from_o->getAddress(false));
 455             } else {
 456                 $from_name = _("Unknown sender");
 457             }
 458             $description = $from_name;
 459         } else {
 460             $default_page = SM_PATH . 'src/download.php';
 461             if (is_object($header->disposition)) {
 462                 $filename = $header->disposition->getProperty('filename');
 463                 if (trim($filename) == '') {
 464                     $name = decodeHeader($header->disposition->getProperty('name'));
 465                     if (trim($name) == '') {
 466                         $name = $header->getParameter('name');
 467                         if(trim($name) == '') {
 468                             if (trim( $header->id ) == '') {
 469                                 $filename = 'untitled-[' . $ent . ']' ;
 470                             } else {
 471                                 $filename = 'cid: ' . $header->id;
 472                             }
 473                         } else {
 474                             $filename = $name;
 475                         }
 476                     } else {
 477                         $filename = $name;
 478                     }
 479                 }
 480             } else {
 481                 $filename = $header->getParameter('name');
 482                 if (!trim($filename)) {
 483                     if (trim( $header->id ) == '') {
 484                         $filename = 'untitled-[' . $ent . ']' ;
 485                     } else {
 486                         $filename = 'cid: ' . $header->id;
 487                     }
 488                 }
 489             }
 490             if ($header->description) {
 491                 $description = decodeHeader($header->description);
 492             } else {
 493                 $description = '';
 494             }
 495         }
 496
 497         $display_filename = $filename;
 498         if (isset($passed_ent_id)) {
 499             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 500         } else {
 501             $passed_ent_id_link = '';
 502         }
 503         $defaultlink = $default_page . "?startMessage=$startMessage"
 504                      . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 505                      . '&amp;ent_id='.$ent.$passed_ent_id_link;
 506         if ($where && $what) {
 507            $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 508         }
 509         /* This executes the attachment hook with a specific MIME-type.
 510          * If that doesn't have results, it tries if there's a rule
 511          * for a more generic type.
 512          */
 513         $hookresults = do_hook("attachment $type0/$type1", $links,
 514                                $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 515                                $display_filename, $where, $what);
 516         if(count($hookresults[1]) <= 1) {
 517             $hookresults = do_hook("attachment $type0/*", $links,
 518                                    $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 519                                    $display_filename, $where, $what);
 520         }
 521
 522         $links = $hookresults[1];
 523         $defaultlink = $hookresults[6];
 524
 525         $attachments .= '<TR><TD>' .
 526                         '<A HREF="'.$defaultlink.'">'.decodeHeader($display_filename).'</A>&nbsp;</TD>' .
 527                         '<TD><SMALL><b>' . show_readable_size($header->size) .
 528                         '</b>&nbsp;&nbsp;</small></TD>' .
 529                         "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
 530                         '<TD><SMALL>';
 531         $attachments .= '<b>' . $description . '</b>';
 532         $attachments .= '</SMALL></TD><TD><SMALL>&nbsp;';
 533
 534         $skipspaces = 1;
 535         foreach ($links as $val) {
 536             if ($skipspaces) {
 537                 $skipspaces = 0;
 538             } else {
 539                 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
 540             }
 541             $attachments .= '<a href="' . $val['href'] . '">' .  $val['text'] . '</a>';
 542         }
 543         unset($links);
 544         $attachments .= "</TD></TR>\n";
 545     }
 546     $attachmentadd = do_hook_function('attachments_bottom',$attachments);
 547     if ($attachmentadd != '')
 548         $attachments = $attachmentadd;
 549     return $attachments;
 550 }
 551
 552 function sqimap_base64_decode(&$string) {
 553     $string = str_replace("\r\n", "\n", $string);
 554     $string = base64_decode($string);
 555 }
 556
 557 /* This function decodes the body depending on the encoding type. */
 558 function decodeBody($body, $encoding) {
 559     global $show_html_default;
 560
 561     $body = str_replace("\r\n", "\n", $body);
 562     $encoding = strtolower($encoding);
 563
 564     $encoding_handler = do_hook_function('decode_body', $encoding);
 565
 566
 567     // plugins get first shot at decoding the body
 568     //
 569     if (!empty($encoding_handler) && function_exists($encoding_handler)) {
 570         $body = $encoding_handler('decode', $body);
 571
 572     } else if ($encoding == 'quoted-printable' ||
 573         $encoding == 'quoted_printable') {
 574         $body = quoted_printable_decode($body);
 575
 576         while (ereg("=\n", $body)) {
 577             $body = ereg_replace ("=\n", '', $body);
 578         }
 579
 580     } else if ($encoding == 'base64') {
 581         $body = base64_decode($body);
 582     }
 583
 584     // All other encodings are returned raw.
 585     return $body;
 586 }
 587
 588 /*
 589  * This functions decode strings that is encoded according to
 590  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 591  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 592  */
 593 function decodeHeader ($string, $utfencode=true,$htmlsave=true) {
 594     global $languages, $squirrelmail_language;
 595     if (is_array($string)) {
 596         $string = implode("\n", $string);
 597     }
 598
 599     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 600         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 601         $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
 602         // Do we need to return at this point?
 603         // return $string;
 604     }
 605     $i = 0;
 606     $iLastMatch = -2;
 607     $encoded = true;
 608
 609     $aString = explode(' ',$string);
 610     $ret = '';
 611     foreach ($aString as $chunk) {
 612         if ($encoded && $chunk === '') {
 613             continue;
 614         } elseif ($chunk === '') {
 615             $ret .= ' ';
 616             continue;
 617         }
 618         $encoded = false;
 619         /* if encoded words are not separated by a linear-space-white we still catch them */
 620         $j = $i-1;
 621
 622         while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 623             /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
 624             if ($iLastMatch !== $j) {
 625                 if ($htmlsave) {
 626                     $ret .= '&nbsp;';
 627                 } else {
 628                     $ret .= ' ';
 629                 }
 630             }
 631             $iLastMatch = $i;
 632             $j = $i;
 633             $ret .= $res[1];
 634             $encoding = ucfirst($res[3]);
 635             switch ($encoding)
 636             {
 637             case 'B':
 638                 $replace = base64_decode($res[4]);
 639                 $ret .= charset_decode($res[2],$replace);
 640                 break;
 641             case 'Q':
 642                 $replace = str_replace('_', ' ', $res[4]);
 643                 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 644                                     $replace);
 645                 /* Only encode into entities by default. Some places
 646                  * don't need the encoding, like the compose form.
 647                  */
 648                 if ($utfencode) {
 649                     $replace = charset_decode($res[2], $replace);
 650                 } else {
 651                     if ($htmlsave) {
 652                         $replace = htmlspecialchars($replace);
 653                     }
 654                 }
 655                 $ret .= $replace;
 656                 break;
 657             default:
 658                 break;
 659             }
 660             $chunk = $res[5];
 661             $encoded = true;
 662         }
 663         if (!$encoded) {
 664             if ($htmlsave) {
 665                 $ret .= '&nbsp;';
 666             } else {
 667                 $ret .= ' ';
 668             }
 669         }
 670
 671         if (!$encoded && $htmlsave) {
 672             $ret .= htmlspecialchars($chunk);
 673         } else {
 674             $ret .= $chunk;
 675         }
 676         ++$i;
 677     }
 678     /* remove the first added space */
 679     if ($ret) {
 680         if ($htmlsave) {
 681             $ret = substr($ret,6);
 682         } else {
 683             $ret = substr($ret,1);
 684         }
 685     }
 686
 687     return $ret;
 688 }
 689
 690 /*
 691  * Encode a string according to RFC 1522 for use in headers if it
 692  * contains 8-bit characters or anything that looks like it should
 693  * be encoded.
 694  */
 695 function encodeHeader ($string) {
 696     global $default_charset, $languages, $squirrelmail_language;
 697
 698     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 699         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 700         return  $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
 701     }
 702     if (strtolower($default_charset) == 'iso-8859-1') {
 703         $string = str_replace("\240",' ',$string);
 704     }
 705
 706     // Encode only if the string contains 8-bit characters or =?
 707     $j = strlen($string);
 708     $max_l = 75 - strlen($default_charset) - 7;
 709     $aRet = array();
 710     $ret = '';
 711     $iEncStart = $enc_init = false;
 712     $cur_l = $iOffset = 0;
 713     for($i = 0; $i < $j; ++$i) {
 714         switch($string{$i})
 715         {
 716         case '=':
 717         case '<':
 718         case '>':
 719         case ',':
 720         case '?':
 721         case '_':
 722             if ($iEncStart === false) {
 723                 $iEncStart = $i;
 724             }
 725             $cur_l+=3;
 726             if ($cur_l > ($max_l-2)) {
 727                 /* if there is an stringpart that doesn't need encoding, add it */
 728                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 729                 $aRet[] = "=?$default_charset?Q?$ret?=";
 730                 $iOffset = $i;
 731                 $cur_l = 0;
 732                 $ret = '';
 733                 $iEncStart = false;
 734             } else {
 735                 $ret .= sprintf("=%02X",ord($string{$i}));
 736             }
 737             break;
 738         case '(':
 739         case ')':
 740             if ($iEncStart !== false) {
 741                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 742                 $aRet[] = "=?$default_charset?Q?$ret?=";
 743                 $iOffset = $i;
 744                 $cur_l = 0;
 745                 $ret = '';
 746                 $iEncStart = false;
 747             }
 748             break;
 749         case ' ':
 750             if ($iEncStart !== false) {
 751                 $cur_l++;
 752                 if ($cur_l > $max_l) {
 753                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 754                     $aRet[] = "=?$default_charset?Q?$ret?=";
 755                     $iOffset = $i;
 756                     $cur_l = 0;
 757                     $ret = '';
 758                     $iEncStart = false;
 759                 } else {
 760                     $ret .= '_';
 761                 }
 762             }
 763             break;
 764         default:
 765             $k = ord($string{$i});
 766             if ($k > 126) {
 767                 if ($iEncStart === false) {
 768                     // do not start encoding in the middle of a string, also take the rest of the word.
 769                     $sLeadString = substr($string,0,$i);
 770                     $aLeadString = explode(' ',$sLeadString);
 771                     $sToBeEncoded = array_pop($aLeadString);
 772                     $iEncStart = $i - strlen($sToBeEncoded);
 773                     $ret .= $sToBeEncoded;
 774                     $cur_l += strlen($sToBeEncoded);
 775                 }
 776                 $cur_l += 3;
 777                 /* first we add the encoded string that reached it's max size */
 778                 if ($cur_l > ($max_l-2)) {
 779                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 780                     $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
 781                     $cur_l = 3;
 782                     $ret = '';
 783                     $iOffset = $i;
 784                     $iEncStart = $i;
 785                 }
 786                 $enc_init = true;
 787                 $ret .= sprintf("=%02X", $k);
 788             } else {
 789                 if ($iEncStart !== false) {
 790                     $cur_l++;
 791                     if ($cur_l > $max_l) {
 792                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 793                         $aRet[] = "=?$default_charset?Q?$ret?=";
 794                         $iEncStart = false;
 795                         $iOffset = $i;
 796                         $cur_l = 0;
 797                         $ret = '';
 798                     } else {
 799                         $ret .= $string{$i};
 800                     }
 801                 }
 802             }
 803             break;
 804         }
 805     }
 806
 807     if ($enc_init) {
 808         if ($iEncStart !== false) {
 809             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 810             $aRet[] = "=?$default_charset?Q?$ret?=";
 811         } else {
 812             $aRet[] = substr($string,$iOffset);
 813         }
 814         $string = implode('',$aRet);
 815     }
 816     return $string;
 817 }
 818
 819 /* This function trys to locate the entity_id of a specific mime element */
 820 function find_ent_id($id, $message) {
 821     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
 822         if ($message->entities[$i]->header->type0 == 'multipart')  {
 823             $ret = find_ent_id($id, $message->entities[$i]);
 824         } else {
 825             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
 826 //                if (sq_check_save_extension($message->entities[$i])) {
 827                     return $message->entities[$i]->entity_id;
 828 //                }
 829             }
 830         }
 831     }
 832     return $ret;
 833 }
 834
 835 function sq_check_save_extension($message) {
 836     $filename = $message->getFilename();
 837     $ext = substr($filename, strrpos($filename,'.')+1);
 838     $save_extensions = array('jpg','jpeg','gif','png','bmp');
 839     return in_array($ext, $save_extensions);
 840 }
 841
 842
 843 /**
 844  ** HTMLFILTER ROUTINES
 845  */
 846
 847 /**
 848  * This function is more or less a wrapper around stripslashes. Apparently
 849  * Explorer is stupid enough to just remove the backslashes and then
 850  * execute the content of the attribute as if nothing happened.
 851  * Who does that?
 852  *
 853  * @param  attvalue   The value of the attribute
 854  * @return attvalue   The value of the attribute stripslashed.
 855  */
 856 function sq_unbackslash($attvalue){
 857     /**
 858      * Remove any backslashes. See if there are any first.
 859      */
 860
 861     if (strstr($attvalue, '\\') !== false){
 862         $attvalue = stripslashes($attvalue);
 863     }
 864     return $attvalue;
 865 }
 866
 867 /**
 868  * Kill any tabs, newlines, or carriage returns. Our friends the
 869  * makers of the browser with 95% market value decided that it'd
 870  * be funny to make "java[tab]script" be just as good as "javascript".
 871  *
 872  * @param  attvalue  The attribute value before extraneous spaces removed.
 873  * @return attvalue  The attribute value after extraneous spaces removed.
 874  */
 875 function sq_unspace($attvalue){
 876     if (strcspn($attvalue, "\t\r\n") != strlen($attvalue)){
 877         $attvalue = str_replace(Array("\t", "\r", "\n"), Array('', '', ''),
 878                                 $attvalue);
 879     }
 880     return $attvalue;
 881 }
 882
 883 /**
 884  * This function returns the final tag out of the tag name, an array
 885  * of attributes, and the type of the tag. This function is called by
 886  * sq_sanitize internally.
 887  *
 888  * @param  $tagname  the name of the tag.
 889  * @param  $attary   the array of attributes and their values
 890  * @param  $tagtype  The type of the tag (see in comments).
 891  * @return           a string with the final tag representation.
 892  */
 893 function sq_tagprint($tagname, $attary, $tagtype){
 894     $me = 'sq_tagprint';
 895
 896     if ($tagtype == 2){
 897         $fulltag = '</' . $tagname . '>';
 898     } else {
 899         $fulltag = '<' . $tagname;
 900         if (is_array($attary) && sizeof($attary)){
 901             $atts = Array();
 902             while (list($attname, $attvalue) = each($attary)){
 903                 array_push($atts, "$attname=$attvalue");
 904             }
 905             $fulltag .= ' ' . join(" ", $atts);
 906         }
 907         if ($tagtype == 3){
 908             $fulltag .= ' /';
 909         }
 910         $fulltag .= '>';
 911     }
 912     return $fulltag;
 913 }
 914
 915 /**
 916  * A small helper function to use with array_walk. Modifies a by-ref
 917  * value and makes it lowercase.
 918  *
 919  * @param  $val a value passed by-ref.
 920  * @return      void since it modifies a by-ref value.
 921  */
 922 function sq_casenormalize(&$val){
 923     $val = strtolower($val);
 924 }
 925
 926 /**
 927  * This function skips any whitespace from the current position within
 928  * a string and to the next non-whitespace value.
 929  *
 930  * @param  $body   the string
 931  * @param  $offset the offset within the string where we should start
 932  *                 looking for the next non-whitespace character.
 933  * @return         the location within the $body where the next
 934  *                 non-whitespace char is located.
 935  */
 936 function sq_skipspace($body, $offset){
 937     $me = 'sq_skipspace';
 938     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
 939     if (sizeof($matches{1})){
 940         $count = strlen($matches{1});
 941         $offset += $count;
 942     }
 943     return $offset;
 944 }
 945
 946 /**
 947  * This function looks for the next character within a string.  It's
 948  * really just a glorified "strpos", except it catches if failures
 949  * nicely.
 950  *
 951  * @param  $body   The string to look for needle in.
 952  * @param  $offset Start looking from this position.
 953  * @param  $needle The character/string to look for.
 954  * @return         location of the next occurance of the needle, or
 955  *                 strlen($body) if needle wasn't found.
 956  */
 957 function sq_findnxstr($body, $offset, $needle){
 958     $me  = 'sq_findnxstr';
 959     $pos = strpos($body, $needle, $offset);
 960     if ($pos === FALSE){
 961         $pos = strlen($body);
 962     }
 963     return $pos;
 964 }
 965
 966 /**
 967  * This function takes a PCRE-style regexp and tries to match it
 968  * within the string.
 969  *
 970  * @param  $body   The string to look for needle in.
 971  * @param  $offset Start looking from here.
 972  * @param  $reg    A PCRE-style regex to match.
 973  * @return         Returns a false if no matches found, or an array
 974  *                 with the following members:
 975  *                 - integer with the location of the match within $body
 976  *                 - string with whatever content between offset and the match
 977  *                 - string with whatever it is we matched
 978  */
 979 function sq_findnxreg($body, $offset, $reg){
 980     $me = 'sq_findnxreg';
 981     $matches = Array();
 982     $retarr = Array();
 983     preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
 984     if (!isset($matches{0}) || !$matches{0}){
 985         $retarr = false;
 986     } else {
 987         $retarr{0} = $offset + strlen($matches{1});
 988         $retarr{1} = $matches{1};
 989         $retarr{2} = $matches{2};
 990     }
 991     return $retarr;
 992 }
 993
 994 /**
 995  * This function looks for the next tag.
 996  *
 997  * @param  $body   String where to look for the next tag.
 998  * @param  $offset Start looking from here.
 999  * @return         false if no more tags exist in the body, or
1000  *                 an array with the following members:
1001  *                 - string with the name of the tag
1002  *                 - array with attributes and their values
1003  *                 - integer with tag type (1, 2, or 3)
1004  *                 - integer where the tag starts (starting "<")
1005  *                 - integer where the tag ends (ending ">")
1006  *                 first three members will be false, if the tag is invalid.
1007  */
1008 function sq_getnxtag($body, $offset){
1009     $me = 'sq_getnxtag';
1010     if ($offset > strlen($body)){
1011         return false;
1012     }
1013     $lt = sq_findnxstr($body, $offset, "<");
1014     if ($lt == strlen($body)){
1015         return false;
1016     }
1017     /**
1018      * We are here:
1019      * blah blah <tag attribute="value">
1020      * \---------^
1021      */
1022     $pos = sq_skipspace($body, $lt+1);
1023     if ($pos >= strlen($body)){
1024         return Array(false, false, false, $lt, strlen($body));
1025     }
1026     /**
1027      * There are 3 kinds of tags:
1028      * 1. Opening tag, e.g.:
1029      *    <a href="blah">
1030      * 2. Closing tag, e.g.:
1031      *    </a>
1032      * 3. XHTML-style content-less tag, e.g.:
1033      *    <img src="blah"/>
1034      */
1035     $tagtype = false;
1036     switch (substr($body, $pos, 1)){
1037         case '/':
1038             $tagtype = 2;
1039             $pos++;
1040             break;
1041         case '!':
1042             /**
1043              * A comment or an SGML declaration.
1044              */
1045             if (substr($body, $pos+1, 2) == "--"){
1046                 $gt = strpos($body, "-->", $pos);
1047                 if ($gt === false){
1048                     $gt = strlen($body);
1049                 } else {
1050                     $gt += 2;
1051                 }
1052                 return Array(false, false, false, $lt, $gt);
1053             } else {
1054                 $gt = sq_findnxstr($body, $pos, ">");
1055                 return Array(false, false, false, $lt, $gt);
1056             }
1057             break;
1058         default:
1059             /**
1060              * Assume tagtype 1 for now. If it's type 3, we'll switch values
1061              * later.
1062              */
1063             $tagtype = 1;
1064             break;
1065     }
1066
1067     $tag_start = $pos;
1068     $tagname = '';
1069     /**
1070      * Look for next [\W-_], which will indicate the end of the tag name.
1071      */
1072     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1073     if ($regary == false){
1074         return Array(false, false, false, $lt, strlen($body));
1075     }
1076     list($pos, $tagname, $match) = $regary;
1077     $tagname = strtolower($tagname);
1078
1079     /**
1080      * $match can be either of these:
1081      * '>'  indicating the end of the tag entirely.
1082      * '\s' indicating the end of the tag name.
1083      * '/'  indicating that this is type-3 xhtml tag.
1084      *
1085      * Whatever else we find there indicates an invalid tag.
1086      */
1087     switch ($match){
1088         case '/':
1089             /**
1090              * This is an xhtml-style tag with a closing / at the
1091              * end, like so: <img src="blah"/>. Check if it's followed
1092              * by the closing bracket. If not, then this tag is invalid
1093              */
1094             if (substr($body, $pos, 2) == "/>"){
1095                 $pos++;
1096                 $tagtype = 3;
1097             } else {
1098                 $gt = sq_findnxstr($body, $pos, ">");
1099                 $retary = Array(false, false, false, $lt, $gt);
1100                 return $retary;
1101             }
1102         case '>':
1103             return Array($tagname, false, $tagtype, $lt, $pos);
1104             break;
1105         default:
1106             /**
1107              * Check if it's whitespace
1108              */
1109             if (!preg_match('/\s/', $match)){
1110                 /**
1111                  * This is an invalid tag! Look for the next closing ">".
1112                  */
1113                 $gt = sq_findnxstr($body, $lt, ">");
1114                 return Array(false, false, false, $lt, $gt);
1115             }
1116             break;
1117     }
1118
1119     /**
1120      * At this point we're here:
1121      * <tagname  attribute='blah'>
1122      * \-------^
1123      *
1124      * At this point we loop in order to find all attributes.
1125      */
1126     $attname = '';
1127     $atttype = false;
1128     $attary = Array();
1129
1130     while ($pos <= strlen($body)){
1131         $pos = sq_skipspace($body, $pos);
1132         if ($pos == strlen($body)){
1133             /**
1134              * Non-closed tag.
1135              */
1136             return Array(false, false, false, $lt, $pos);
1137         }
1138         /**
1139          * See if we arrived at a ">" or "/>", which means that we reached
1140          * the end of the tag.
1141          */
1142         $matches = Array();
1143         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1144             /**
1145              * Yep. So we did.
1146              */
1147             $pos += strlen($matches{1});
1148             if ($matches{2} == "/>"){
1149                 $tagtype = 3;
1150                 $pos++;
1151             }
1152             return Array($tagname, $attary, $tagtype, $lt, $pos);
1153         }
1154
1155         /**
1156          * There are several types of attributes, with optional
1157          * [:space:] between members.
1158          * Type 1:
1159          *   attrname[:space:]=[:space:]'CDATA'
1160          * Type 2:
1161          *   attrname[:space:]=[:space:]"CDATA"
1162          * Type 3:
1163          *   attr[:space:]=[:space:]CDATA
1164          * Type 4:
1165          *   attrname
1166          *
1167          * We leave types 1 and 2 the same, type 3 we check for
1168          * '"' and convert to "&quot" if needed, then wrap in
1169          * double quotes. Type 4 we convert into:
1170          * attrname="yes".
1171          */
1172         $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1173         if ($regary == false){
1174             /**
1175              * Looks like body ended before the end of tag.
1176              */
1177             return Array(false, false, false, $lt, strlen($body));
1178         }
1179         list($pos, $attname, $match) = $regary;
1180         $attname = strtolower($attname);
1181         /**
1182          * We arrived at the end of attribute name. Several things possible
1183          * here:
1184          * '>'  means the end of the tag and this is attribute type 4
1185          * '/'  if followed by '>' means the same thing as above
1186          * '\s' means a lot of things -- look what it's followed by.
1187          *      anything else means the attribute is invalid.
1188          */
1189         switch($match){
1190             case '/':
1191                 /**
1192                  * This is an xhtml-style tag with a closing / at the
1193                  * end, like so: <img src="blah"/>. Check if it's followed
1194                  * by the closing bracket. If not, then this tag is invalid
1195                  */
1196                 if (substr($body, $pos, 2) == "/>"){
1197                     $pos++;
1198                     $tagtype = 3;
1199                 } else {
1200                     $gt = sq_findnxstr($body, $pos, ">");
1201                     $retary = Array(false, false, false, $lt, $gt);
1202                     return $retary;
1203                 }
1204             case '>':
1205                 $attary{$attname} = '"yes"';
1206                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1207                 break;
1208             default:
1209                 /**
1210                  * Skip whitespace and see what we arrive at.
1211                  */
1212                 $pos = sq_skipspace($body, $pos);
1213                 $char = substr($body, $pos, 1);
1214                 /**
1215                  * Two things are valid here:
1216                  * '=' means this is attribute type 1 2 or 3.
1217                  * \w means this was attribute type 4.
1218                  * anything else we ignore and re-loop. End of tag and
1219                  * invalid stuff will be caught by our checks at the beginning
1220                  * of the loop.
1221                  */
1222                 if ($char == "="){
1223                     $pos++;
1224                     $pos = sq_skipspace($body, $pos);
1225                     /**
1226                      * Here are 3 possibilities:
1227                      * "'"  attribute type 1
1228                      * '"'  attribute type 2
1229                      * everything else is the content of tag type 3
1230                      */
1231                     $quot = substr($body, $pos, 1);
1232                     if ($quot == "'"){
1233                         $regary = sq_findnxreg($body, $pos+1, "\'");
1234                         if ($regary == false){
1235                             return Array(false, false, false, $lt, strlen($body));
1236                         }
1237                         list($pos, $attval, $match) = $regary;
1238                         $pos++;
1239                         $attary{$attname} = "'" . $attval . "'";
1240                     } else if ($quot == '"'){
1241                         $regary = sq_findnxreg($body, $pos+1, '\"');
1242                         if ($regary == false){
1243                             return Array(false, false, false, $lt, strlen($body));
1244                         }
1245                         list($pos, $attval, $match) = $regary;
1246                         $pos++;
1247                         $attary{$attname} = '"' . $attval . '"';
1248                     } else {
1249                         /**
1250                          * These are hateful. Look for \s, or >.
1251                          */
1252                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1253                         if ($regary == false){
1254                             return Array(false, false, false, $lt, strlen($body));
1255                         }
1256                         list($pos, $attval, $match) = $regary;
1257                         /**
1258                          * If it's ">" it will be caught at the top.
1259                          */
1260                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1261                         $attary{$attname} = '"' . $attval . '"';
1262                     }
1263                 } else if (preg_match("|[\w/>]|", $char)) {
1264                     /**
1265                      * That was attribute type 4.
1266                      */
1267                     $attary{$attname} = '"yes"';
1268                 } else {
1269                     /**
1270                      * An illegal character. Find next '>' and return.
1271                      */
1272                     $gt = sq_findnxstr($body, $pos, ">");
1273                     return Array(false, false, false, $lt, $gt);
1274                 }
1275                 break;
1276         }
1277     }
1278     /**
1279      * The fact that we got here indicates that the tag end was never
1280      * found. Return invalid tag indication so it gets stripped.
1281      */
1282     return Array(false, false, false, $lt, strlen($body));
1283 }
1284
1285 /**
1286  * This function checks attribute values for entity-encoded values
1287  * and returns them translated into 8-bit strings so we can run
1288  * checks on them.
1289  *
1290  * @param  $attvalue A string to run entity check against.
1291  * @return           Translated value.
1292  */
1293
1294 function sq_deent($attvalue){
1295     $me = 'sq_deent';
1296     /**
1297      * See if we have to run the checks first. All entities must start
1298      * with "&".
1299      */
1300     if (strpos($attvalue, '&') === false){
1301         return $attvalue;
1302     }
1303     /**
1304      * Check named entities first.
1305      */
1306     $trans = get_html_translation_table(HTML_ENTITIES);
1307     /**
1308      * Leave &quot; in, as it can mess us up.
1309      */
1310     $trans = array_flip($trans);
1311     unset($trans{'&quot;'});
1312     while (list($ent, $val) = each($trans)){
1313         $attvalue = preg_replace('/' . $ent . '*/si', $val, $attvalue);
1314     }
1315     /**
1316      * Now translate numbered entities from 1 to 255 if needed.
1317      */
1318     if (strpos($attvalue, '#') !== false){
1319         $omit = Array(34, 39);
1320         for ($asc = 256; $asc >= 0; $asc--){
1321             if (!in_array($asc, $omit)){
1322                 $chr = chr($asc);
1323                 $octrule = '/\&#0*' . $asc . ';*/si';
1324                 $hexrule = '/\&#x0*' . dechex($asc) . ';*/si';
1325                 $attvalue = preg_replace($octrule, $chr, $attvalue);
1326                 $attvalue = preg_replace($hexrule, $chr, $attvalue);
1327             }
1328         }
1329     }
1330     return $attvalue;
1331 }
1332
1333 /**
1334  * This function runs various checks against the attributes.
1335  *
1336  * @param  $tagname         String with the name of the tag.
1337  * @param  $attary          Array with all tag attributes.
1338  * @param  $rm_attnames     See description for sq_sanitize
1339  * @param  $bad_attvals     See description for sq_sanitize
1340  * @param  $add_attr_to_tag See description for sq_sanitize
1341  * @param  $message         message object
1342  * @param  $id              message id
1343  * @return                  Array with modified attributes.
1344  */
1345 function sq_fixatts($tagname,
1346                     $attary,
1347                     $rm_attnames,
1348                     $bad_attvals,
1349                     $add_attr_to_tag,
1350                     $message,
1351                     $id,
1352                     $mailbox
1353                     ){
1354     $me = 'sq_fixatts';
1355     while (list($attname, $attvalue) = each($attary)){
1356         /**
1357          * See if this attribute should be removed.
1358          */
1359         foreach ($rm_attnames as $matchtag=>$matchattrs){
1360             if (preg_match($matchtag, $tagname)){
1361                 foreach ($matchattrs as $matchattr){
1362                     if (preg_match($matchattr, $attname)){
1363                         unset($attary{$attname});
1364                         continue;
1365                     }
1366                 }
1367             }
1368         }
1369         /**
1370          * Remove any backslashes, entities, and extraneous whitespace.
1371          */
1372         $attvalue = sq_unbackslash($attvalue);
1373         $attvalue = sq_deent($attvalue);
1374         $attvalue = sq_unspace($attvalue);
1375
1376         /**
1377          * Now let's run checks on the attvalues.
1378          * I don't expect anyone to comprehend this. If you do,
1379          * get in touch with me so I can drive to where you live and
1380          * shake your hand personally. :)
1381          */
1382         foreach ($bad_attvals as $matchtag=>$matchattrs){
1383             if (preg_match($matchtag, $tagname)){
1384                 foreach ($matchattrs as $matchattr=>$valary){
1385                     if (preg_match($matchattr, $attname)){
1386                         /**
1387                          * There are two arrays in valary.
1388                          * First is matches.
1389                          * Second one is replacements
1390                          */
1391                         list($valmatch, $valrepl) = $valary;
1392                         $newvalue =
1393                             preg_replace($valmatch, $valrepl, $attvalue);
1394                         if ($newvalue != $attvalue){
1395                             $attary{$attname} = $newvalue;
1396                         }
1397                     }
1398                 }
1399             }
1400         }
1401         /**
1402          * Turn cid: urls into http-friendly ones.
1403          */
1404         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1405             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1406         }
1407     }
1408     /**
1409      * See if we need to append any attributes to this tag.
1410      */
1411     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1412         if (preg_match($matchtag, $tagname)){
1413             $attary = array_merge($attary, $addattary);
1414         }
1415     }
1416     return $attary;
1417 }
1418
1419 /**
1420  * This function edits the style definition to make them friendly and
1421  * usable in squirrelmail.
1422  *
1423  * @param  $message  the message object
1424  * @param  $id       the message id
1425  * @param  $content  a string with whatever is between <style> and </style>
1426  * @param  $mailbox  the message mailbox
1427  * @return           a string with edited content.
1428  */
1429 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1430     global $view_unsafe_images;
1431     $me = 'sq_fixstyle';
1432     $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1433     if ($ret == FALSE){
1434         return array(FALSE, strlen($body));
1435     }
1436     $newpos = $ret[0] + strlen($ret[2]);
1437     $content = $ret[1];
1438     /**
1439      * First look for general BODY style declaration, which would be
1440      * like so:
1441      * body {background: blah-blah}
1442      * and change it to .bodyclass so we can just assign it to a <div>
1443      */
1444     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1445     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1446     /**
1447      * Fix url('blah') declarations.
1448      */
1449     $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1450                             "url(\\1$secremoveimg\\2)", $content);
1451     /**
1452      * Fix url('https*://.*) declarations but only if $view_unsafe_images
1453      * is false.
1454      */
1455     if (!$view_unsafe_images){
1456         $content = preg_replace("|url\s*\(\s*([\'\"])\s*https*:.*?([\'\"])\s*\)|si",
1457                                 "url(\\1$secremoveimg\\2)", $content);
1458     }
1459
1460     /**
1461      * Fix urls that refer to cid:
1462      */
1463     while (preg_match("|url\s*\(\s*([\'\"]\s*cid:.*?[\'\"])\s*\)|si",
1464                       $content, $matches)){
1465         $cidurl = $matches{1};
1466         $httpurl = sq_cid2http($message, $id, $cidurl, $mailbox);
1467         $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1468                                 "url($httpurl)", $content);
1469     }
1470
1471     /**
1472      * Fix stupid css declarations which lead to vulnerabilities
1473      * in IE.
1474      */
1475     $match   = Array('/expression/i',
1476                      '/behaviou*r/i',
1477                      '/binding/i',
1478                      '/include-source/i');
1479     $replace = Array('idiocy', 'idiocy', 'idiocy', 'idiocy');
1480     $content = preg_replace($match, $replace, $content);
1481     return array($content, $newpos);
1482 }
1483
1484 /**
1485  * This function converts cid: url's into the ones that can be viewed in
1486  * the browser.
1487  *
1488  * @param  $message  the message object
1489  * @param  $id       the message id
1490  * @param  $cidurl   the cid: url.
1491  * @param  $mailbox  the message mailbox
1492  * @return           a string with a http-friendly url
1493  */
1494 function sq_cid2http($message, $id, $cidurl, $mailbox){
1495     /**
1496      * Get rid of quotes.
1497      */
1498     $quotchar = substr($cidurl, 0, 1);
1499     if ($quotchar == '"' || $quotchar == "'"){
1500         $cidurl = str_replace($quotchar, "", $cidurl);
1501     } else {
1502         $quotchar = '';
1503     }
1504     $cidurl = substr(trim($cidurl), 4);
1505     $linkurl = find_ent_id($cidurl, $message);
1506     /* in case of non-save cid links $httpurl should be replaced by a sort of
1507        unsave link image */
1508     $httpurl = '';
1509     if ($linkurl) {
1510         $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1511                    "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1512                    '&amp;ent_id=' . $linkurl . $quotchar;
1513     }
1514     return $httpurl;
1515 }
1516
1517 /**
1518  * This function changes the <body> tag into a <div> tag since we
1519  * can't really have a body-within-body.
1520  *
1521  * @param  $attary   an array of attributes and values of <body>
1522  * @param  $mailbox  mailbox we're currently reading (for cid2http)
1523  * @param  $message  current message (for cid2http)
1524  * @param  $id       current message id (for cid2http)
1525  * @return           a modified array of attributes to be set for <div>
1526  */
1527 function sq_body2div($attary, $mailbox, $message, $id){
1528     $me = 'sq_body2div';
1529     $divattary = Array('class' => "'bodyclass'");
1530     $bgcolor = '#ffffff';
1531     $text = '#000000';
1532     $has_bgc_stl = $has_txt_stl = false;
1533     $styledef = '';
1534     if (is_array($attary) && sizeof($attary) > 0){
1535         foreach ($attary as $attname=>$attvalue){
1536             $quotchar = substr($attvalue, 0, 1);
1537             $attvalue = str_replace($quotchar, "", $attvalue);
1538             switch ($attname){
1539                 case 'background':
1540                     $attvalue = sq_cid2http($message, $id,
1541                                             $attvalue, $mailbox);
1542                     $styledef .= "background-image: url('$attvalue'); ";
1543                     break;
1544                 case 'bgcolor':
1545                     $has_bgc_stl = true;
1546                     $styledef .= "background-color: $attvalue; ";
1547                     break;
1548                 case 'text':
1549                     $has_txt_stl = true;
1550                     $styledef .= "color: $attvalue; ";
1551                     break;
1552             }
1553         }
1554         // Outlook defines a white bgcolor and no text color. This can lead to
1555         // white text on a white bg with certain themes.
1556         if ($has_bgc_stl && !$has_txt_stl) {
1557             $styledef .= "color: $text; ";
1558         }
1559         if (strlen($styledef) > 0){
1560             $divattary{"style"} = "\"$styledef\"";
1561         }
1562     }
1563     return $divattary;
1564 }
1565
1566 /**
1567  * This is the main function and the one you should actually be calling.
1568  * There are several variables you should be aware of an which need
1569  * special description.
1570  *
1571  * Since the description is quite lengthy, see it here:
1572  * http://www.mricon.com/html/phpfilter.html
1573  *
1574  * @param $body                 the string with HTML you wish to filter
1575  * @param $tag_list             see description above
1576  * @param $rm_tags_with_content see description above
1577  * @param $self_closing_tags    see description above
1578  * @param $force_tag_closing    see description above
1579  * @param $rm_attnames          see description above
1580  * @param $bad_attvals          see description above
1581  * @param $add_attr_to_tag      see description above
1582  * @param $message              message object
1583  * @param $id                   message id
1584  * @return                      sanitized html safe to show on your pages.
1585  */
1586 function sq_sanitize($body,
1587                      $tag_list,
1588                      $rm_tags_with_content,
1589                      $self_closing_tags,
1590                      $force_tag_closing,
1591                      $rm_attnames,
1592                      $bad_attvals,
1593                      $add_attr_to_tag,
1594                      $message,
1595                      $id,
1596                      $mailbox
1597                      ){
1598     $me = 'sq_sanitize';
1599     $rm_tags = array_shift($tag_list);
1600     /**
1601      * Normalize rm_tags and rm_tags_with_content.
1602      */
1603     @array_walk($tag_list, 'sq_casenormalize');
1604     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1605     @array_walk($self_closing_tags, 'sq_casenormalize');
1606     /**
1607      * See if tag_list is of tags to remove or tags to allow.
1608      * false  means remove these tags
1609      * true   means allow these tags
1610      */
1611     $curpos = 0;
1612     $open_tags = Array();
1613     $trusted = "\n<!-- begin sanitized html -->\n";
1614     $skip_content = false;
1615     /**
1616      * Take care of netscape's stupid javascript entities like
1617      * &{alert('boo')};
1618      */
1619     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1620
1621     while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1622         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1623         $free_content = substr($body, $curpos, $lt-$curpos);
1624         /**
1625          * Take care of <style>
1626          */
1627         if ($tagname == "style" && $tagtype == 1){
1628             list($free_content, $curpos) =
1629                 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
1630             if ($free_content != FALSE){
1631                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1632                 $trusted .= $free_content;
1633                 $trusted .= sq_tagprint($tagname, false, 2);
1634             }
1635             continue;
1636         }
1637         if ($skip_content == false){
1638             $trusted .= $free_content;
1639         }
1640         if ($tagname != FALSE){
1641             if ($tagtype == 2){
1642                 if ($skip_content == $tagname){
1643                     /**
1644                      * Got to the end of tag we needed to remove.
1645                      */
1646                     $tagname = false;
1647                     $skip_content = false;
1648                 } else {
1649                     if ($skip_content == false){
1650                         if ($tagname == "body"){
1651                             $tagname = "div";
1652                         }
1653                         if (isset($open_tags{$tagname}) &&
1654                             $open_tags{$tagname} > 0){
1655                             $open_tags{$tagname}--;
1656                         } else {
1657                             $tagname = false;
1658                         }
1659                     }
1660                 }
1661             } else {
1662                 /**
1663                  * $rm_tags_with_content
1664                  */
1665                 if ($skip_content == false){
1666                     /**
1667                      * See if this is a self-closing type and change
1668                      * tagtype appropriately.
1669                      */
1670                     if ($tagtype == 1
1671                         && in_array($tagname, $self_closing_tags)){
1672                         $tagtype = 3;
1673                     }
1674                     /**
1675                      * See if we should skip this tag and any content
1676                      * inside it.
1677                      */
1678                     if ($tagtype == 1 &&
1679                         in_array($tagname, $rm_tags_with_content)){
1680                         $skip_content = $tagname;
1681                     } else {
1682                         if (($rm_tags == false
1683                              && in_array($tagname, $tag_list)) ||
1684                             ($rm_tags == true &&
1685                              !in_array($tagname, $tag_list))){
1686                             $tagname = false;
1687                         } else {
1688                             /**
1689                              * Convert body into div.
1690                              */
1691                             if ($tagname == "body"){
1692                                 $tagname = "div";
1693                                 $attary = sq_body2div($attary, $mailbox,
1694                                                       $message, $id);
1695                             }
1696                             if ($tagtype == 1){
1697                                 if (isset($open_tags{$tagname})){
1698                                     $open_tags{$tagname}++;
1699                                 } else {
1700                                     $open_tags{$tagname}=1;
1701                                 }
1702                             }
1703                             /**
1704                              * This is where we run other checks.
1705                              */
1706                             if (is_array($attary) && sizeof($attary) > 0){
1707                                 $attary = sq_fixatts($tagname,
1708                                                      $attary,
1709                                                      $rm_attnames,
1710                                                      $bad_attvals,
1711                                                      $add_attr_to_tag,
1712                                                      $message,
1713                                                      $id,
1714                                                      $mailbox
1715                                                      );
1716                             }
1717                         }
1718                     }
1719                 }
1720             }
1721             if ($tagname != false && $skip_content == false){
1722                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1723             }
1724         }
1725         $curpos = $gt+1;
1726     }
1727     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1728     if ($force_tag_closing == true){
1729         foreach ($open_tags as $tagname=>$opentimes){
1730             while ($opentimes > 0){
1731                 $trusted .= '</' . $tagname . '>';
1732                 $opentimes--;
1733             }
1734         }
1735         $trusted .= "\n";
1736     }
1737     $trusted .= "<!-- end sanitized html -->\n";
1738     return $trusted;
1739 }
1740
1741 /**
1742  * This is a wrapper function to call html sanitizing routines.
1743  *
1744  * @param  $body  the body of the message
1745  * @param  $id    the id of the message
1746  * @return        a string with html safe to display in the browser.
1747  */
1748 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1749     global $attachment_common_show_images, $view_unsafe_images,
1750            $has_unsafe_images;
1751     /**
1752      * Don't display attached images in HTML mode.
1753      */
1754     $attachment_common_show_images = false;
1755     $tag_list = Array(
1756                       false,
1757                       "object",
1758                       "meta",
1759                       "html",
1760                       "head",
1761                       "base",
1762                       "link",
1763                       "frame",
1764                       "iframe",
1765                       "plaintext",
1766                       "marquee"
1767                       );
1768
1769     $rm_tags_with_content = Array(
1770                                   "script",
1771                                   "applet",
1772                                   "embed",
1773                                   "title",
1774                                   "frameset",
1775                                   "xml"
1776                                   );
1777
1778     $self_closing_tags =  Array(
1779                                 "img",
1780                                 "br",
1781                                 "hr",
1782                                 "input"
1783                                 );
1784
1785     $force_tag_closing = true;
1786
1787     $rm_attnames = Array(
1788                          "/.*/" =>
1789                          Array(
1790                                "/target/i",
1791                                "/^on.*/i",
1792                                "/^dynsrc/i",
1793                                "/^data.*/i",
1794                                "/^lowsrc.*/i"
1795                                )
1796                          );
1797
1798     $secremoveimg = "../images/" . _("sec_remove_eng.png");
1799     $bad_attvals = Array(
1800         "/.*/" =>
1801             Array(
1802                 "/^src|background/i" =>
1803                     Array(
1804                           Array(
1805                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1806                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1807                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1808                                 ),
1809                           Array(
1810                                 "\\1$secremoveimg\\2",
1811                                 "\\1$secremoveimg\\2",
1812                                 "\\1$secremoveimg\\2",
1813                                 "\\1$secremoveimg\\2"
1814                                 )
1815                         ),
1816                 "/^href|action/i" =>
1817                     Array(
1818                           Array(
1819                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1820                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1821                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1822                                 ),
1823                           Array(
1824                                 "\\1#\\1",
1825                                 "\\1#\\1",
1826                                 "\\1#\\1",
1827                                 "\\1#\\1"
1828                                 )
1829                         ),
1830                 "/^style/i" =>
1831                     Array(
1832                           Array(
1833                                 "/expression/i",
1834                                 "/binding/i",
1835                                 "/behaviou*r/i",
1836                                 "/include-source/i",
1837                                 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
1838                                 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
1839                                 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si"
1840                                ),
1841                           Array(
1842                                 "idiocy",
1843                                 "idiocy",
1844                                 "idiocy",
1845                                 "idiocy",
1846                                 "url(\\1#\\1)",
1847                                 "url(\\1#\\1)",
1848                                 "url(\\1#\\1)",
1849                                 "url(\\1#\\1)"
1850                                )
1851                           )
1852                 )
1853         );
1854     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1855         $view_unsafe_images = false;
1856     }
1857     if (!$view_unsafe_images){
1858         /**
1859          * Remove any references to http/https if view_unsafe_images set
1860          * to false.
1861          */
1862          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1863                     '/^([\'\"])\s*https*:.*([\'\"])/si');
1864          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1865                     "\\1$secremoveimg\\1");
1866          array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
1867                     '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1868          array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
1869                     "url(\\1$secremoveimg\\1)");
1870     }
1871
1872     $add_attr_to_tag = Array(
1873         "/^a$/i" =>
1874             Array('target'=>'"_new"',
1875                   'title'=>'"'._("This external link will open in a new window").'"'
1876             )
1877     );
1878     $trusted = sq_sanitize($body,
1879                            $tag_list,
1880                            $rm_tags_with_content,
1881                            $self_closing_tags,
1882                            $force_tag_closing,
1883                            $rm_attnames,
1884                            $bad_attvals,
1885                            $add_attr_to_tag,
1886                            $message,
1887                            $id,
1888                            $mailbox
1889                            );
1890     if (preg_match("|$secremoveimg|i", $trusted)){
1891         $has_unsafe_images = true;
1892     }
1893     return $trusted;
1894 }
1895
1896 /**
1897  * function SendDownloadHeaders - send file to the browser
1898  *
1899  * Original Source: SM core src/download.php
1900  * moved here to make it available to other code, and separate
1901  * front end from back end functionality.
1902  *
1903  * @param string $type0 first half of mime type
1904  * @param string $type1 second half of mime type
1905  * @param string $filename filename to tell the browser for downloaded file
1906  * @param boolean $force whether to force the download dialog to pop
1907  * @param integer $filesize optional, send the Content-Header and length to the browser
1908  * @return void
1909  */
1910  function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
1911      global $languages, $squirrelmail_language;
1912      $isIE = $isIE6 = 0;
1913
1914      sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
1915
1916      if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
1917          strstr($HTTP_USER_AGENT, 'Opera') === false) {
1918          $isIE = 1;
1919      }
1920
1921      if (strstr($HTTP_USER_AGENT, 'compatible; MSIE 6') !== false &&
1922          strstr($HTTP_USER_AGENT, 'Opera') === false) {
1923          $isIE6 = 1;
1924      }
1925
1926      if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
1927          function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
1928          $filename =
1929              $languages[$squirrelmail_language]['XTRA_CODE']('downloadfilename', $filename, $HTTP_USER_AGENT);
1930      } else {
1931         $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
1932      }
1933
1934      // A Pox on Microsoft and it's Office!
1935      if (!$force) {
1936          // Try to show in browser window
1937          header("Content-Disposition: inline; filename=\"$filename\"");
1938          header("Content-Type: $type0/$type1; name=\"$filename\"");
1939      } else {
1940          // Try to pop up the "save as" box
1941          // IE makes this hard.  It pops up 2 save boxes, or none.
1942          // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
1943          // But, accordint to Microsoft, it is "RFC compliant but doesn't
1944          // take into account some deviations that allowed within the
1945          // specification."  Doesn't that mean RFC non-compliant?
1946          // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
1947          //
1948          // The best thing you can do for IE is to upgrade to the latest
1949          // version
1950          if ($isIE && !$isIE6) {
1951              // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
1952              // Do not have quotes around filename, but that applied to
1953              // "attachment"... does it apply to inline too?
1954              //
1955              // This combination seems to work mostly.  IE 5.5 SP 1 has
1956              // known issues (see the Microsoft Knowledge Base)
1957              header("Content-Disposition: inline; filename=$filename");
1958              // This works for most types, but doesn't work with Word files
1959              header("Content-Type: application/download; name=\"$filename\"");
1960
1961              // These are spares, just in case.  :-)
1962              //header("Content-Type: $type0/$type1; name=\"$filename\"");
1963              //header("Content-Type: application/x-msdownload; name=\"$filename\"");
1964              //header("Content-Type: application/octet-stream; name=\"$filename\"");
1965          } else {
1966              header("Content-Disposition: attachment; filename=\"$filename\"");
1967              // application/octet-stream forces download for Netscape
1968              header("Content-Type: application/octet-stream; name=\"$filename\"");
1969          }
1970      }
1971
1972      //send the content-length header if the calling function provides it
1973      if ($filesize > 0) {
1974         header("Content-Length: $filesize");
1975      }
1976
1977  }  // end fn SendDownlaodHeaders
1978
1979 ?>