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 = 'read_body.php?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         if ($view_unsafe_images) {
 403             $text = _("Hide Unsafe Images");
 404         } else {
 405             if (isset($has_unsafe_images) && $has_unsafe_images) {
 406                 $link .= '&amp;view_unsafe_images=1';
 407                 $text = _("View Unsafe Images");
 408             } else {
 409                 $text = '';
 410             }
 411         }
 412         $body .= '<center><small><a href="'.$link.'">'.$text.
 413                  '</a></small></center><br>' . "\n";
 414     }
 415     return $body;
 416 }
 417
 418
 419 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 420     global $where, $what, $startMessage, $color;
 421     static $ShownHTML = 0;
 422
 423     $att_ar = $message->getAttachments($exclude_id);
 424
 425     if (!count($att_ar)) return '';
 426
 427     $attachments = '';
 428
 429     $urlMailbox = urlencode($mailbox);
 430
 431     foreach ($att_ar as $att) {
 432         $ent = $att->entity_id;
 433         $header = $att->header;
 434         $type0 = strtolower($header->type0);
 435         $type1 = strtolower($header->type1);
 436         $name = '';
 437         $links['download link']['text'] = _("download");
 438         $links['download link']['href'] = SM_PATH .
 439                 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 440         $ImageURL = '';
 441         if ($type0 =='message' && $type1 == 'rfc822') {
 442             $default_page = SM_PATH . 'src/read_body.php';
 443             $rfc822_header = $att->rfc822_header;
 444             $filename = $rfc822_header->subject;
 445             if (trim( $filename ) == '') {
 446                 $filename = 'untitled-[' . $ent . ']' ;
 447             }
 448             $from_o = $rfc822_header->from;
 449             if (is_object($from_o)) {
 450                 $from_name = decodeHeader($from_o->getAddress(false));
 451             } else {
 452                 $from_name = _("Unknown sender");
 453             }
 454             $description = $from_name;
 455         } else {
 456             $default_page = SM_PATH . 'src/download.php';
 457             if (is_object($header->disposition)) {
 458                 $filename = $header->disposition->getProperty('filename');
 459                 if (trim($filename) == '') {
 460                     $name = decodeHeader($header->disposition->getProperty('name'));
 461                     if (trim($name) == '') {
 462                         $name = $header->getParameter('name');
 463                         if(trim($name) == '') {
 464                             if (trim( $header->id ) == '') {
 465                                 $filename = 'untitled-[' . $ent . ']' ;
 466                             } else {
 467                                 $filename = 'cid: ' . $header->id;
 468                             }
 469                         } else {
 470                             $filename = $name;
 471                         }
 472                     } else {
 473                         $filename = $name;
 474                     }
 475                 }
 476             } else {
 477                 $filename = $header->getParameter('name');
 478                 if (!trim($filename)) {
 479                     if (trim( $header->id ) == '') {
 480                         $filename = 'untitled-[' . $ent . ']' ;
 481                     } else {
 482                         $filename = 'cid: ' . $header->id;
 483                     }
 484                 }
 485             }
 486             if ($header->description) {
 487                 $description = decodeHeader($header->description);
 488             } else {
 489                 $description = '';
 490             }
 491         }
 492
 493         $display_filename = $filename;
 494         if (isset($passed_ent_id)) {
 495             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 496         } else {
 497             $passed_ent_id_link = '';
 498         }
 499         $defaultlink = $default_page . "?startMessage=$startMessage"
 500                      . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 501                      . '&amp;ent_id='.$ent.$passed_ent_id_link;
 502         if ($where && $what) {
 503            $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 504         }
 505         /* This executes the attachment hook with a specific MIME-type.
 506          * If that doesn't have results, it tries if there's a rule
 507          * for a more generic type.
 508          */
 509         $hookresults = do_hook("attachment $type0/$type1", $links,
 510                                $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 511                                $display_filename, $where, $what);
 512         if(count($hookresults[1]) <= 1) {
 513             $hookresults = do_hook("attachment $type0/*", $links,
 514                                    $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 515                                    $display_filename, $where, $what);
 516         }
 517
 518         $links = $hookresults[1];
 519         $defaultlink = $hookresults[6];
 520
 521         $attachments .= '<TR><TD>' .
 522                         '<A HREF="'.$defaultlink.'">'.decodeHeader($display_filename).'</A>&nbsp;</TD>' .
 523                         '<TD><SMALL><b>' . show_readable_size($header->size) .
 524                         '</b>&nbsp;&nbsp;</small></TD>' .
 525                         "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
 526                         '<TD><SMALL>';
 527         $attachments .= '<b>' . $description . '</b>';
 528         $attachments .= '</SMALL></TD><TD><SMALL>&nbsp;';
 529
 530         $skipspaces = 1;
 531         foreach ($links as $val) {
 532             if ($skipspaces) {
 533                 $skipspaces = 0;
 534             } else {
 535                 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
 536             }
 537             $attachments .= '<a href="' . $val['href'] . '">' .  $val['text'] . '</a>';
 538         }
 539         unset($links);
 540         $attachments .= "</TD></TR>\n";
 541     }
 542     $attachmentadd = do_hook_function('attachments_bottom',$attachments);
 543     if ($attachmentadd != '')
 544         $attachments = $attachmentadd;
 545     return $attachments;
 546 }
 547
 548 function sqimap_base64_decode(&$string) {
 549     $string = str_replace("\r\n", "\n", $string);
 550     $string = base64_decode($string);
 551 }
 552
 553 /* This function decodes the body depending on the encoding type. */
 554 function decodeBody($body, $encoding) {
 555     global $show_html_default;
 556
 557     $body = str_replace("\r\n", "\n", $body);
 558     $encoding = strtolower($encoding);
 559
 560     $encoding_handler = do_hook_function('decode_body', $encoding);
 561
 562
 563     // plugins get first shot at decoding the body
 564     //
 565     if (!empty($encoding_handler) && function_exists($encoding_handler)) {
 566         $body = $encoding_handler('decode', $body);
 567
 568     } else if ($encoding == 'quoted-printable' ||
 569         $encoding == 'quoted_printable') {
 570         $body = quoted_printable_decode($body);
 571
 572         while (ereg("=\n", $body)) {
 573             $body = ereg_replace ("=\n", '', $body);
 574         }
 575
 576     } else if ($encoding == 'base64') {
 577         $body = base64_decode($body);
 578     }
 579
 580     // All other encodings are returned raw.
 581     return $body;
 582 }
 583
 584 /*
 585  * This functions decode strings that is encoded according to
 586  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 587  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 588  */
 589 function decodeHeader ($string, $utfencode=true,$htmlsave=true) {
 590     global $languages, $squirrelmail_language;
 591     if (is_array($string)) {
 592         $string = implode("\n", $string);
 593     }
 594
 595     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 596         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 597         $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
 598         // Do we need to return at this point?
 599         // return $string;
 600     }
 601     $i = 0;
 602     $iLastMatch = -2;
 603     $encoded = true;
 604
 605     $aString = explode(' ',$string);
 606     $ret = '';
 607     foreach ($aString as $chunk) {
 608         if ($encoded && $chunk === '') {
 609             continue;
 610         } elseif ($chunk === '') {
 611             $ret .= ' ';
 612             continue;
 613         }
 614         $encoded = false;
 615         /* if encoded words are not separated by a linear-space-white we still catch them */
 616         $j = $i-1;
 617
 618         while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 619             /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
 620             if ($iLastMatch !== $j) {
 621                 if ($htmlsave) {
 622                     $ret .= '&nbsp;';
 623                 } else {
 624                     $ret .= ' ';
 625                 }
 626             }
 627             $iLastMatch = $i;
 628             $j = $i;
 629             $ret .= $res[1];
 630             $encoding = ucfirst($res[3]);
 631             switch ($encoding)
 632             {
 633             case 'B':
 634                 $replace = base64_decode($res[4]);
 635                 $ret .= charset_decode($res[2],$replace);
 636                 break;
 637             case 'Q':
 638                 $replace = str_replace('_', ' ', $res[4]);
 639                 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 640                                     $replace);
 641                 /* Only encode into entities by default. Some places
 642                  * don't need the encoding, like the compose form.
 643                  */
 644                 if ($utfencode) {
 645                     $replace = charset_decode($res[2], $replace);
 646                 } else {
 647                     if ($htmlsave) {
 648                         $replace = htmlspecialchars($replace);
 649                     }
 650                 }
 651                 $ret .= $replace;
 652                 break;
 653             default:
 654                 break;
 655             }
 656             $chunk = $res[5];
 657             $encoded = true;
 658         }
 659
 660         if (!$encoded && $htmlsave) {
 661             $ret .= htmlspecialchars($chunk);
 662         } else {
 663             $ret .= $chunk;
 664         }
 665
 666         if (!$encoded) {
 667             if ($htmlsave) {
 668                 $ret .= '&nbsp;';
 669             } else {
 670                 $ret .= ' ';
 671             }
 672         }
 673         ++$i;
 674     }
 675
 676     return $ret;
 677 }
 678
 679 /*
 680  * Encode a string according to RFC 1522 for use in headers if it
 681  * contains 8-bit characters or anything that looks like it should
 682  * be encoded.
 683  */
 684 function encodeHeader ($string) {
 685     global $default_charset, $languages, $squirrelmail_language;
 686
 687     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 688         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 689         return  $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
 690     }
 691     if (strtolower($default_charset) == 'iso-8859-1') {
 692         $string = str_replace("\240",' ',$string);
 693     }
 694
 695     // Encode only if the string contains 8-bit characters or =?
 696     $j = strlen($string);
 697     $max_l = 75 - strlen($default_charset) - 7;
 698     $aRet = array();
 699     $ret = '';
 700     $iEncStart = $enc_init = false;
 701     $cur_l = $iOffset = 0;
 702     for($i = 0; $i < $j; ++$i) {
 703         switch($string{$i})
 704         {
 705         case '=':
 706         case '<':
 707         case '>':
 708         case ',':
 709         case '?':
 710         case '_':
 711             if ($iEncStart === false) {
 712                 $iEncStart = $i;
 713             }
 714             $cur_l+=3;
 715             if ($cur_l > ($max_l-2)) {
 716                 /* if there is an stringpart that doesn't need encoding, add it */
 717                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 718                 $aRet[] = "=?$default_charset?Q?$ret?=";
 719                 $iOffset = $i;
 720                 $cur_l = 0;
 721                 $ret = '';
 722                 $iEncStart = false;
 723             } else {
 724                 $ret .= sprintf("=%02X",ord($string{$i}));
 725             }
 726             break;
 727         case '(':
 728         case ')':
 729             if ($iEncStart !== false) {
 730                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 731                 $aRet[] = "=?$default_charset?Q?$ret?=";
 732                 $iOffset = $i;
 733                 $cur_l = 0;
 734                 $ret = '';
 735                 $iEncStart = false;
 736             }
 737             break;
 738         case ' ':
 739             if ($iEncStart !== false) {
 740                 $cur_l++;
 741                 if ($cur_l > $max_l) {
 742                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 743                     $aRet[] = "=?$default_charset?Q?$ret?=";
 744                     $iOffset = $i;
 745                     $cur_l = 0;
 746                     $ret = '';
 747                     $iEncStart = false;
 748                 } else {
 749                     $ret .= '_';
 750                 }
 751             }
 752             break;
 753         default:
 754             $k = ord($string{$i});
 755             if ($k > 126) {
 756                 if ($iEncStart === false) {
 757                     // do not start encoding in the middle of a string, also take the rest of the word.
 758                     $sLeadString = substr($string,0,$i);
 759                     $aLeadString = explode(' ',$sLeadString);
 760                     $sToBeEncoded = array_pop($aLeadString);
 761                     $iEncStart = $i - strlen($sToBeEncoded);
 762                     $ret .= $sToBeEncoded;
 763                     $cur_l += strlen($sToBeEncoded);
 764                 }
 765                 $cur_l += 3;
 766                 /* first we add the encoded string that reached it's max size */
 767                 if ($cur_l > ($max_l-2)) {
 768                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 769                     $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
 770                     $cur_l = 3;
 771                     $ret = '';
 772                     $iOffset = $i;
 773                     $iEncStart = $i;
 774                 }
 775                 $enc_init = true;
 776                 $ret .= sprintf("=%02X", $k);
 777             } else {
 778                 if ($iEncStart !== false) {
 779                     $cur_l++;
 780                     if ($cur_l > $max_l) {
 781                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 782                         $aRet[] = "=?$default_charset?Q?$ret?=";
 783                         $iEncStart = false;
 784                         $iOffset = $i;
 785                         $cur_l = 0;
 786                         $ret = '';
 787                     } else {
 788                         $ret .= $string{$i};
 789                     }
 790                 }
 791             }
 792             break;
 793         }
 794     }
 795
 796     if ($enc_init) {
 797         if ($iEncStart !== false) {
 798             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 799             $aRet[] = "=?$default_charset?Q?$ret?=";
 800         } else {
 801             $aRet[] = substr($string,$iOffset);
 802         }
 803         $string = implode('',$aRet);
 804     }
 805     return $string;
 806 }
 807
 808 /* This function trys to locate the entity_id of a specific mime element */
 809 function find_ent_id($id, $message) {
 810     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
 811         if ($message->entities[$i]->header->type0 == 'multipart')  {
 812             $ret = find_ent_id($id, $message->entities[$i]);
 813         } else {
 814             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
 815 //                if (sq_check_save_extension($message->entities[$i])) {
 816                     return $message->entities[$i]->entity_id;
 817 //                }
 818             }
 819         }
 820     }
 821     return $ret;
 822 }
 823
 824 function sq_check_save_extension($message) {
 825     $filename = $message->getFilename();
 826     $ext = substr($filename, strrpos($filename,'.')+1);
 827     $save_extensions = array('jpg','jpeg','gif','png','bmp');
 828     return in_array($ext, $save_extensions);
 829 }
 830
 831
 832 /**
 833  ** HTMLFILTER ROUTINES
 834  */
 835
 836 /**
 837  * This function is more or less a wrapper around stripslashes. Apparently
 838  * Explorer is stupid enough to just remove the backslashes and then
 839  * execute the content of the attribute as if nothing happened.
 840  * Who does that?
 841  *
 842  * @param  attvalue   The value of the attribute
 843  * @return attvalue   The value of the attribute stripslashed.
 844  */
 845 function sq_unbackslash($attvalue){
 846     /**
 847      * Remove any backslashes. See if there are any first.
 848      */
 849     if (strstr($attvalue, '\\') !== false){
 850         $attvalue = stripslashes($attvalue);
 851     }
 852     return $attvalue;
 853 }
 854
 855 /**
 856  * Kill any tabs, newlines, or carriage returns. Our friends the
 857  * makers of the browser with 95% market value decided that it'd
 858  * be funny to make "java[tab]script" be just as good as "javascript".
 859  *
 860  * @param  attvalue  The attribute value before extraneous spaces removed.
 861  * @return attvalue  The attribute value after extraneous spaces removed.
 862  */
 863 function sq_unspace($attvalue){
 864     if (strcspn($attvalue, "\t\r\n") != strlen($attvalue)){
 865         $attvalue = str_replace(Array("\t", "\r", "\n"), Array('', '', ''),
 866                                 $attvalue);
 867     }
 868     return $attvalue;
 869 }
 870
 871 /**
 872  * This function returns the final tag out of the tag name, an array
 873  * of attributes, and the type of the tag. This function is called by
 874  * sq_sanitize internally.
 875  *
 876  * @param  $tagname  the name of the tag.
 877  * @param  $attary   the array of attributes and their values
 878  * @param  $tagtype  The type of the tag (see in comments).
 879  * @return           a string with the final tag representation.
 880  */
 881 function sq_tagprint($tagname, $attary, $tagtype){
 882     $me = 'sq_tagprint';
 883
 884     if ($tagtype == 2){
 885         $fulltag = '</' . $tagname . '>';
 886     } else {
 887         $fulltag = '<' . $tagname;
 888         if (is_array($attary) && sizeof($attary)){
 889             $atts = Array();
 890             while (list($attname, $attvalue) = each($attary)){
 891                 array_push($atts, "$attname=$attvalue");
 892             }
 893             $fulltag .= ' ' . join(" ", $atts);
 894         }
 895         if ($tagtype == 3){
 896             $fulltag .= ' /';
 897         }
 898         $fulltag .= '>';
 899     }
 900     return $fulltag;
 901 }
 902
 903 /**
 904  * A small helper function to use with array_walk. Modifies a by-ref
 905  * value and makes it lowercase.
 906  *
 907  * @param  $val a value passed by-ref.
 908  * @return      void since it modifies a by-ref value.
 909  */
 910 function sq_casenormalize(&$val){
 911     $val = strtolower($val);
 912 }
 913
 914 /**
 915  * This function skips any whitespace from the current position within
 916  * a string and to the next non-whitespace value.
 917  *
 918  * @param  $body   the string
 919  * @param  $offset the offset within the string where we should start
 920  *                 looking for the next non-whitespace character.
 921  * @return         the location within the $body where the next
 922  *                 non-whitespace char is located.
 923  */
 924 function sq_skipspace($body, $offset){
 925     $me = 'sq_skipspace';
 926     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
 927     if (sizeof($matches{1})){
 928         $count = strlen($matches{1});
 929         $offset += $count;
 930     }
 931     return $offset;
 932 }
 933
 934 /**
 935  * This function looks for the next character within a string.  It's
 936  * really just a glorified "strpos", except it catches if failures
 937  * nicely.
 938  *
 939  * @param  $body   The string to look for needle in.
 940  * @param  $offset Start looking from this position.
 941  * @param  $needle The character/string to look for.
 942  * @return         location of the next occurance of the needle, or
 943  *                 strlen($body) if needle wasn't found.
 944  */
 945 function sq_findnxstr($body, $offset, $needle){
 946     $me  = 'sq_findnxstr';
 947     $pos = strpos($body, $needle, $offset);
 948     if ($pos === FALSE){
 949         $pos = strlen($body);
 950     }
 951     return $pos;
 952 }
 953
 954 /**
 955  * This function takes a PCRE-style regexp and tries to match it
 956  * within the string.
 957  *
 958  * @param  $body   The string to look for needle in.
 959  * @param  $offset Start looking from here.
 960  * @param  $reg    A PCRE-style regex to match.
 961  * @return         Returns a false if no matches found, or an array
 962  *                 with the following members:
 963  *                 - integer with the location of the match within $body
 964  *                 - string with whatever content between offset and the match
 965  *                 - string with whatever it is we matched
 966  */
 967 function sq_findnxreg($body, $offset, $reg){
 968     $me = 'sq_findnxreg';
 969     $matches = Array();
 970     $retarr = Array();
 971     preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
 972     if (!isset($matches{0}) || !$matches{0}){
 973         $retarr = false;
 974     } else {
 975         $retarr{0} = $offset + strlen($matches{1});
 976         $retarr{1} = $matches{1};
 977         $retarr{2} = $matches{2};
 978     }
 979     return $retarr;
 980 }
 981
 982 /**
 983  * This function looks for the next tag.
 984  *
 985  * @param  $body   String where to look for the next tag.
 986  * @param  $offset Start looking from here.
 987  * @return         false if no more tags exist in the body, or
 988  *                 an array with the following members:
 989  *                 - string with the name of the tag
 990  *                 - array with attributes and their values
 991  *                 - integer with tag type (1, 2, or 3)
 992  *                 - integer where the tag starts (starting "<")
 993  *                 - integer where the tag ends (ending ">")
 994  *                 first three members will be false, if the tag is invalid.
 995  */
 996 function sq_getnxtag($body, $offset){
 997     $me = 'sq_getnxtag';
 998     if ($offset > strlen($body)){
 999         return false;
1000     }
1001     $lt = sq_findnxstr($body, $offset, "<");
1002     if ($lt == strlen($body)){
1003         return false;
1004     }
1005     /**
1006      * We are here:
1007      * blah blah <tag attribute="value">
1008      * \---------^
1009      */
1010     $pos = sq_skipspace($body, $lt+1);
1011     if ($pos >= strlen($body)){
1012         return Array(false, false, false, $lt, strlen($body));
1013     }
1014     /**
1015      * There are 3 kinds of tags:
1016      * 1. Opening tag, e.g.:
1017      *    <a href="blah">
1018      * 2. Closing tag, e.g.:
1019      *    </a>
1020      * 3. XHTML-style content-less tag, e.g.:
1021      *    <img src="blah"/>
1022      */
1023     $tagtype = false;
1024     switch (substr($body, $pos, 1)){
1025         case '/':
1026             $tagtype = 2;
1027             $pos++;
1028             break;
1029         case '!':
1030             /**
1031              * A comment or an SGML declaration.
1032              */
1033             if (substr($body, $pos+1, 2) == "--"){
1034                 $gt = strpos($body, "-->", $pos);
1035                 if ($gt === false){
1036                     $gt = strlen($body);
1037                 } else {
1038                     $gt += 2;
1039                 }
1040                 return Array(false, false, false, $lt, $gt);
1041             } else {
1042                 $gt = sq_findnxstr($body, $pos, ">");
1043                 return Array(false, false, false, $lt, $gt);
1044             }
1045             break;
1046         default:
1047             /**
1048              * Assume tagtype 1 for now. If it's type 3, we'll switch values
1049              * later.
1050              */
1051             $tagtype = 1;
1052             break;
1053     }
1054
1055     $tag_start = $pos;
1056     $tagname = '';
1057     /**
1058      * Look for next [\W-_], which will indicate the end of the tag name.
1059      */
1060     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1061     if ($regary == false){
1062         return Array(false, false, false, $lt, strlen($body));
1063     }
1064     list($pos, $tagname, $match) = $regary;
1065     $tagname = strtolower($tagname);
1066
1067     /**
1068      * $match can be either of these:
1069      * '>'  indicating the end of the tag entirely.
1070      * '\s' indicating the end of the tag name.
1071      * '/'  indicating that this is type-3 xhtml tag.
1072      *
1073      * Whatever else we find there indicates an invalid tag.
1074      */
1075     switch ($match){
1076         case '/':
1077             /**
1078              * This is an xhtml-style tag with a closing / at the
1079              * end, like so: <img src="blah"/>. Check if it's followed
1080              * by the closing bracket. If not, then this tag is invalid
1081              */
1082             if (substr($body, $pos, 2) == "/>"){
1083                 $pos++;
1084                 $tagtype = 3;
1085             } else {
1086                 $gt = sq_findnxstr($body, $pos, ">");
1087                 $retary = Array(false, false, false, $lt, $gt);
1088                 return $retary;
1089             }
1090         case '>':
1091             return Array($tagname, false, $tagtype, $lt, $pos);
1092             break;
1093         default:
1094             /**
1095              * Check if it's whitespace
1096              */
1097             if (!preg_match('/\s/', $match)){
1098                 /**
1099                  * This is an invalid tag! Look for the next closing ">".
1100                  */
1101                 $gt = sq_findnxstr($body, $lt, ">");
1102                 return Array(false, false, false, $lt, $gt);
1103             }
1104             break;
1105     }
1106
1107     /**
1108      * At this point we're here:
1109      * <tagname  attribute='blah'>
1110      * \-------^
1111      *
1112      * At this point we loop in order to find all attributes.
1113      */
1114     $attname = '';
1115     $atttype = false;
1116     $attary = Array();
1117
1118     while ($pos <= strlen($body)){
1119         $pos = sq_skipspace($body, $pos);
1120         if ($pos == strlen($body)){
1121             /**
1122              * Non-closed tag.
1123              */
1124             return Array(false, false, false, $lt, $pos);
1125         }
1126         /**
1127          * See if we arrived at a ">" or "/>", which means that we reached
1128          * the end of the tag.
1129          */
1130         $matches = Array();
1131         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1132             /**
1133              * Yep. So we did.
1134              */
1135             $pos += strlen($matches{1});
1136             if ($matches{2} == "/>"){
1137                 $tagtype = 3;
1138                 $pos++;
1139             }
1140             return Array($tagname, $attary, $tagtype, $lt, $pos);
1141         }
1142
1143         /**
1144          * There are several types of attributes, with optional
1145          * [:space:] between members.
1146          * Type 1:
1147          *   attrname[:space:]=[:space:]'CDATA'
1148          * Type 2:
1149          *   attrname[:space:]=[:space:]"CDATA"
1150          * Type 3:
1151          *   attr[:space:]=[:space:]CDATA
1152          * Type 4:
1153          *   attrname
1154          *
1155          * We leave types 1 and 2 the same, type 3 we check for
1156          * '"' and convert to "&quot" if needed, then wrap in
1157          * double quotes. Type 4 we convert into:
1158          * attrname="yes".
1159          */
1160         $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1161         if ($regary == false){
1162             /**
1163              * Looks like body ended before the end of tag.
1164              */
1165             return Array(false, false, false, $lt, strlen($body));
1166         }
1167         list($pos, $attname, $match) = $regary;
1168         $attname = strtolower($attname);
1169         /**
1170          * We arrived at the end of attribute name. Several things possible
1171          * here:
1172          * '>'  means the end of the tag and this is attribute type 4
1173          * '/'  if followed by '>' means the same thing as above
1174          * '\s' means a lot of things -- look what it's followed by.
1175          *      anything else means the attribute is invalid.
1176          */
1177         switch($match){
1178             case '/':
1179                 /**
1180                  * This is an xhtml-style tag with a closing / at the
1181                  * end, like so: <img src="blah"/>. Check if it's followed
1182                  * by the closing bracket. If not, then this tag is invalid
1183                  */
1184                 if (substr($body, $pos, 2) == "/>"){
1185                     $pos++;
1186                     $tagtype = 3;
1187                 } else {
1188                     $gt = sq_findnxstr($body, $pos, ">");
1189                     $retary = Array(false, false, false, $lt, $gt);
1190                     return $retary;
1191                 }
1192             case '>':
1193                 $attary{$attname} = '"yes"';
1194                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1195                 break;
1196             default:
1197                 /**
1198                  * Skip whitespace and see what we arrive at.
1199                  */
1200                 $pos = sq_skipspace($body, $pos);
1201                 $char = substr($body, $pos, 1);
1202                 /**
1203                  * Two things are valid here:
1204                  * '=' means this is attribute type 1 2 or 3.
1205                  * \w means this was attribute type 4.
1206                  * anything else we ignore and re-loop. End of tag and
1207                  * invalid stuff will be caught by our checks at the beginning
1208                  * of the loop.
1209                  */
1210                 if ($char == "="){
1211                     $pos++;
1212                     $pos = sq_skipspace($body, $pos);
1213                     /**
1214                      * Here are 3 possibilities:
1215                      * "'"  attribute type 1
1216                      * '"'  attribute type 2
1217                      * everything else is the content of tag type 3
1218                      */
1219                     $quot = substr($body, $pos, 1);
1220                     if ($quot == "'"){
1221                         $regary = sq_findnxreg($body, $pos+1, "\'");
1222                         if ($regary == false){
1223                             return Array(false, false, false, $lt, strlen($body));
1224                         }
1225                         list($pos, $attval, $match) = $regary;
1226                         $pos++;
1227                         $attary{$attname} = "'" . $attval . "'";
1228                     } else if ($quot == '"'){
1229                         $regary = sq_findnxreg($body, $pos+1, '\"');
1230                         if ($regary == false){
1231                             return Array(false, false, false, $lt, strlen($body));
1232                         }
1233                         list($pos, $attval, $match) = $regary;
1234                         $pos++;
1235                         $attary{$attname} = '"' . $attval . '"';
1236                     } else {
1237                         /**
1238                          * These are hateful. Look for \s, or >.
1239                          */
1240                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1241                         if ($regary == false){
1242                             return Array(false, false, false, $lt, strlen($body));
1243                         }
1244                         list($pos, $attval, $match) = $regary;
1245                         /**
1246                          * If it's ">" it will be caught at the top.
1247                          */
1248                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1249                         $attary{$attname} = '"' . $attval . '"';
1250                     }
1251                 } else if (preg_match("|[\w/>]|", $char)) {
1252                     /**
1253                      * That was attribute type 4.
1254                      */
1255                     $attary{$attname} = '"yes"';
1256                 } else {
1257                     /**
1258                      * An illegal character. Find next '>' and return.
1259                      */
1260                     $gt = sq_findnxstr($body, $pos, ">");
1261                     return Array(false, false, false, $lt, $gt);
1262                 }
1263                 break;
1264         }
1265     }
1266     /**
1267      * The fact that we got here indicates that the tag end was never
1268      * found. Return invalid tag indication so it gets stripped.
1269      */
1270     return Array(false, false, false, $lt, strlen($body));
1271 }
1272
1273 /**
1274  * This function checks attribute values for entity-encoded values
1275  * and returns them translated into 8-bit strings so we can run
1276  * checks on them.
1277  *
1278  * @param  $attvalue A string to run entity check against.
1279  * @return           Translated value.
1280  */
1281 function sq_deent($attvalue){
1282     $me = 'sq_deent';
1283     /**
1284      * See if we have to run the checks first. All entities must start
1285      * with "&".
1286      */
1287     if (strpos($attvalue, "&") === false){
1288         return $attvalue;
1289     }
1290     /**
1291      * Check named entities first.
1292      */
1293     $trans = get_html_translation_table(HTML_ENTITIES);
1294     /**
1295      * Leave &quot; in, as it can mess us up.
1296      */
1297     $trans = array_flip($trans);
1298     unset($trans{"&quot;"});
1299     while (list($ent, $val) = each($trans)){
1300         $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1301     }
1302     /**
1303      * Now translate numbered entities from 1 to 255 if needed.
1304      */
1305     if (strpos($attvalue, "#") !== false){
1306         $omit = Array(34, 39);
1307         for ($asc=1; $asc<256; $asc++){
1308             if (!in_array($asc, $omit)){
1309                 $chr = chr($asc);
1310                 $attvalue = preg_replace("/\&#0*$asc;*(\D)/si", "$chr\\1",
1311                                          $attvalue);
1312                 $attvalue = preg_replace("/\&#x0*".dechex($asc).";*(\W)/si",
1313                                          "$chr\\1", $attvalue);
1314             }
1315         }
1316     }
1317     return $attvalue;
1318 }
1319
1320 /**
1321  * This function runs various checks against the attributes.
1322  *
1323  * @param  $tagname         String with the name of the tag.
1324  * @param  $attary          Array with all tag attributes.
1325  * @param  $rm_attnames     See description for sq_sanitize
1326  * @param  $bad_attvals     See description for sq_sanitize
1327  * @param  $add_attr_to_tag See description for sq_sanitize
1328  * @param  $message         message object
1329  * @param  $id              message id
1330  * @return                  Array with modified attributes.
1331  */
1332 function sq_fixatts($tagname,
1333                     $attary,
1334                     $rm_attnames,
1335                     $bad_attvals,
1336                     $add_attr_to_tag,
1337                     $message,
1338                     $id,
1339                     $mailbox
1340                     ){
1341     $me = 'sq_fixatts';
1342     while (list($attname, $attvalue) = each($attary)){
1343         /**
1344          * See if this attribute should be removed.
1345          */
1346         foreach ($rm_attnames as $matchtag=>$matchattrs){
1347             if (preg_match($matchtag, $tagname)){
1348                 foreach ($matchattrs as $matchattr){
1349                     if (preg_match($matchattr, $attname)){
1350                         unset($attary{$attname});
1351                         continue;
1352                     }
1353                 }
1354             }
1355         }
1356         /**
1357          * Remove any backslashes, entities, and extraneous whitespace.
1358          */
1359         $attvalue = sq_unbackslash($attvalue);
1360         $attvalue = sq_deent($attvalue);
1361         $attvalue = sq_unspace($attvalue);
1362
1363         /**
1364          * Now let's run checks on the attvalues.
1365          * I don't expect anyone to comprehend this. If you do,
1366          * get in touch with me so I can drive to where you live and
1367          * shake your hand personally. :)
1368          */
1369         foreach ($bad_attvals as $matchtag=>$matchattrs){
1370             if (preg_match($matchtag, $tagname)){
1371                 foreach ($matchattrs as $matchattr=>$valary){
1372                     if (preg_match($matchattr, $attname)){
1373                         /**
1374                          * There are two arrays in valary.
1375                          * First is matches.
1376                          * Second one is replacements
1377                          */
1378                         list($valmatch, $valrepl) = $valary;
1379                         $newvalue =
1380                             preg_replace($valmatch, $valrepl, $attvalue);
1381                         if ($newvalue != $attvalue){
1382                             $attary{$attname} = $newvalue;
1383                         }
1384                     }
1385                 }
1386             }
1387         }
1388         /**
1389          * Turn cid: urls into http-friendly ones.
1390          */
1391         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1392             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1393         }
1394     }
1395     /**
1396      * See if we need to append any attributes to this tag.
1397      */
1398     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1399         if (preg_match($matchtag, $tagname)){
1400             $attary = array_merge($attary, $addattary);
1401         }
1402     }
1403     return $attary;
1404 }
1405
1406 /**
1407  * This function edits the style definition to make them friendly and
1408  * usable in squirrelmail.
1409  *
1410  * @param  $message  the message object
1411  * @param  $id       the message id
1412  * @param  $content  a string with whatever is between <style> and </style>
1413  * @return           a string with edited content.
1414  */
1415 function sq_fixstyle($body, $pos, $message, $id){
1416     global $view_unsafe_images;
1417     $me = 'sq_fixstyle';
1418     $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1419     if ($ret == FALSE){
1420         return array(FALSE, strlen($body));
1421     }
1422     $newpos = $ret[0] + strlen($ret[2]);
1423     $content = $ret[1];
1424     /**
1425      * First look for general BODY style declaration, which would be
1426      * like so:
1427      * body {background: blah-blah}
1428      * and change it to .bodyclass so we can just assign it to a <div>
1429      */
1430     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1431     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1432     /**
1433      * Fix url('blah') declarations.
1434      */
1435     $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1436                             "url(\\1$secremoveimg\\2)", $content);
1437     /**
1438      * Fix url('https*://.*) declarations but only if $view_unsafe_images
1439      * is false.
1440      */
1441     if (!$view_unsafe_images){
1442         $content = preg_replace("|url\s*\(\s*([\'\"])\s*https*:.*?([\'\"])\s*\)|si",
1443                                 "url(\\1$secremoveimg\\2)", $content);
1444     }
1445
1446     /**
1447      * Fix urls that refer to cid:
1448      */
1449     while (preg_match("|url\s*\(\s*([\'\"]\s*cid:.*?[\'\"])\s*\)|si",
1450                       $content, $matches)){
1451         $cidurl = $matches{1};
1452         $httpurl = sq_cid2http($message, $id, $cidurl);
1453         $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1454                                 "url($httpurl)", $content);
1455     }
1456
1457     /**
1458      * Fix stupid css declarations which lead to vulnerabilities
1459      * in IE.
1460      */
1461     $match   = Array('/expression/i',
1462                      '/behaviou*r/i',
1463                      '/binding/i',
1464                      '/include-source/i');
1465     $replace = Array('idiocy', 'idiocy', 'idiocy', 'idiocy');
1466     $content = preg_replace($match, $replace, $content);
1467     return array($content, $newpos);
1468 }
1469
1470 /**
1471  * This function converts cid: url's into the ones that can be viewed in
1472  * the browser.
1473  *
1474  * @param  $message  the message object
1475  * @param  $id       the message id
1476  * @param  $cidurl   the cid: url.
1477  * @return           a string with a http-friendly url
1478  */
1479 function sq_cid2http($message, $id, $cidurl, $mailbox){
1480     /**
1481      * Get rid of quotes.
1482      */
1483     $quotchar = substr($cidurl, 0, 1);
1484     if ($quotchar == '"' || $quotchar == "'"){
1485         $cidurl = str_replace($quotchar, "", $cidurl);
1486     } else {
1487         $quotchar = '';
1488     }
1489     $cidurl = substr(trim($cidurl), 4);
1490     $linkurl = find_ent_id($cidurl, $message);
1491     /* in case of non-save cid links $httpurl should be replaced by a sort of
1492        unsave link image */
1493     $httpurl = '';
1494     if ($linkurl) {
1495         $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1496                    "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1497                    '&amp;ent_id=' . $linkurl . $quotchar;
1498     }
1499     return $httpurl;
1500 }
1501
1502 /**
1503  * This function changes the <body> tag into a <div> tag since we
1504  * can't really have a body-within-body.
1505  *
1506  * @param  $attary   an array of attributes and values of <body>
1507  * @param  $mailbox  mailbox we're currently reading (for cid2http)
1508  * @param  $message  current message (for cid2http)
1509  * @param  $id       current message id (for cid2http)
1510  * @return           a modified array of attributes to be set for <div>
1511  */
1512 function sq_body2div($attary, $mailbox, $message, $id){
1513     $me = 'sq_body2div';
1514     $divattary = Array('class' => "'bodyclass'");
1515     $bgcolor = '#ffffff';
1516     $text = '#000000';
1517     $has_bgc_stl = $has_txt_stl = false;
1518     $styledef = '';
1519     if (is_array($attary) && sizeof($attary) > 0){
1520         foreach ($attary as $attname=>$attvalue){
1521             $quotchar = substr($attvalue, 0, 1);
1522             $attvalue = str_replace($quotchar, "", $attvalue);
1523             switch ($attname){
1524                 case 'background':
1525                     $attvalue = sq_cid2http($message, $id,
1526                                             $attvalue, $mailbox);
1527                     $styledef .= "background-image: url('$attvalue'); ";
1528                     break;
1529                 case 'bgcolor':
1530                     $has_bgc_stl = true;
1531                     $styledef .= "background-color: $attvalue; ";
1532                     break;
1533                 case 'text':
1534                     $has_txt_stl = true;
1535                     $styledef .= "color: $attvalue; ";
1536                     break;
1537             }
1538         }
1539         // Outlook defines a white bgcolor and no text color. This can lead to
1540         // white text on a white bg with certain themes.
1541         if ($has_bgc_stl && !$has_txt_stl) {
1542             $styledef .= "color: $text; ";
1543         }
1544         if (strlen($styledef) > 0){
1545             $divattary{"style"} = "\"$styledef\"";
1546         }
1547     }
1548     return $divattary;
1549 }
1550
1551 /**
1552  * This is the main function and the one you should actually be calling.
1553  * There are several variables you should be aware of an which need
1554  * special description.
1555  *
1556  * Since the description is quite lengthy, see it here:
1557  * http://www.mricon.com/html/phpfilter.html
1558  *
1559  * @param $body                 the string with HTML you wish to filter
1560  * @param $tag_list             see description above
1561  * @param $rm_tags_with_content see description above
1562  * @param $self_closing_tags    see description above
1563  * @param $force_tag_closing    see description above
1564  * @param $rm_attnames          see description above
1565  * @param $bad_attvals          see description above
1566  * @param $add_attr_to_tag      see description above
1567  * @param $message              message object
1568  * @param $id                   message id
1569  * @return                      sanitized html safe to show on your pages.
1570  */
1571 function sq_sanitize($body,
1572                      $tag_list,
1573                      $rm_tags_with_content,
1574                      $self_closing_tags,
1575                      $force_tag_closing,
1576                      $rm_attnames,
1577                      $bad_attvals,
1578                      $add_attr_to_tag,
1579                      $message,
1580                      $id,
1581                      $mailbox
1582                      ){
1583     $me = 'sq_sanitize';
1584     $rm_tags = array_shift($tag_list);
1585     /**
1586      * Normalize rm_tags and rm_tags_with_content.
1587      */
1588     @array_walk($tag_list, 'sq_casenormalize');
1589     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1590     @array_walk($self_closing_tags, 'sq_casenormalize');
1591     /**
1592      * See if tag_list is of tags to remove or tags to allow.
1593      * false  means remove these tags
1594      * true   means allow these tags
1595      */
1596     $curpos = 0;
1597     $open_tags = Array();
1598     $trusted = "\n<!-- begin sanitized html -->\n";
1599     $skip_content = false;
1600     /**
1601      * Take care of netscape's stupid javascript entities like
1602      * &{alert('boo')};
1603      */
1604     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1605
1606     while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1607         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1608         $free_content = substr($body, $curpos, $lt-$curpos);
1609         /**
1610          * Take care of <style>
1611          */
1612         if ($tagname == "style" && $tagtype == 1){
1613             list($free_content, $curpos) =
1614                 sq_fixstyle($body, $gt+1, $message, $id);
1615             if ($free_content != FALSE){
1616                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1617                 $trusted .= $free_content;
1618                 $trusted .= sq_tagprint($tagname, false, 2);
1619             }
1620             continue;
1621         }
1622         if ($skip_content == false){
1623             $trusted .= $free_content;
1624         }
1625         if ($tagname != FALSE){
1626             if ($tagtype == 2){
1627                 if ($skip_content == $tagname){
1628                     /**
1629                      * Got to the end of tag we needed to remove.
1630                      */
1631                     $tagname = false;
1632                     $skip_content = false;
1633                 } else {
1634                     if ($skip_content == false){
1635                         if ($tagname == "body"){
1636                             $tagname = "div";
1637                         }
1638                         if (isset($open_tags{$tagname}) &&
1639                             $open_tags{$tagname} > 0){
1640                             $open_tags{$tagname}--;
1641                         } else {
1642                             $tagname = false;
1643                         }
1644                     }
1645                 }
1646             } else {
1647                 /**
1648                  * $rm_tags_with_content
1649                  */
1650                 if ($skip_content == false){
1651                     /**
1652                      * See if this is a self-closing type and change
1653                      * tagtype appropriately.
1654                      */
1655                     if ($tagtype == 1
1656                         && in_array($tagname, $self_closing_tags)){
1657                         $tagtype = 3;
1658                     }
1659                     /**
1660                      * See if we should skip this tag and any content
1661                      * inside it.
1662                      */
1663                     if ($tagtype == 1 &&
1664                         in_array($tagname, $rm_tags_with_content)){
1665                         $skip_content = $tagname;
1666                     } else {
1667                         if (($rm_tags == false
1668                              && in_array($tagname, $tag_list)) ||
1669                             ($rm_tags == true &&
1670                              !in_array($tagname, $tag_list))){
1671                             $tagname = false;
1672                         } else {
1673                             /**
1674                              * Convert body into div.
1675                              */
1676                             if ($tagname == "body"){
1677                                 $tagname = "div";
1678                                 $attary = sq_body2div($attary, $mailbox,
1679                                                       $message, $id);
1680                             }
1681                             if ($tagtype == 1){
1682                                 if (isset($open_tags{$tagname})){
1683                                     $open_tags{$tagname}++;
1684                                 } else {
1685                                     $open_tags{$tagname}=1;
1686                                 }
1687                             }
1688                             /**
1689                              * This is where we run other checks.
1690                              */
1691                             if (is_array($attary) && sizeof($attary) > 0){
1692                                 $attary = sq_fixatts($tagname,
1693                                                      $attary,
1694                                                      $rm_attnames,
1695                                                      $bad_attvals,
1696                                                      $add_attr_to_tag,
1697                                                      $message,
1698                                                      $id,
1699                                                      $mailbox
1700                                                      );
1701                             }
1702                         }
1703                     }
1704                 }
1705             }
1706             if ($tagname != false && $skip_content == false){
1707                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1708             }
1709         }
1710         $curpos = $gt+1;
1711     }
1712     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1713     if ($force_tag_closing == true){
1714         foreach ($open_tags as $tagname=>$opentimes){
1715             while ($opentimes > 0){
1716                 $trusted .= '</' . $tagname . '>';
1717                 $opentimes--;
1718             }
1719         }
1720         $trusted .= "\n";
1721     }
1722     $trusted .= "<!-- end sanitized html -->\n";
1723     return $trusted;
1724 }
1725
1726 /**
1727  * This is a wrapper function to call html sanitizing routines.
1728  *
1729  * @param  $body  the body of the message
1730  * @param  $id    the id of the message
1731  * @return        a string with html safe to display in the browser.
1732  */
1733 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1734     global $attachment_common_show_images, $view_unsafe_images,
1735            $has_unsafe_images;
1736     /**
1737      * Don't display attached images in HTML mode.
1738      */
1739     $attachment_common_show_images = false;
1740     $tag_list = Array(
1741                       false,
1742                       "object",
1743                       "meta",
1744                       "html",
1745                       "head",
1746                       "base",
1747                       "link",
1748                       "frame",
1749                       "iframe",
1750                       "plaintext",
1751                       "marquee"
1752                       );
1753
1754     $rm_tags_with_content = Array(
1755                                   "script",
1756                                   "applet",
1757                                   "embed",
1758                                   "title",
1759                                   "frameset",
1760                                   "xml"
1761                                   );
1762
1763     $self_closing_tags =  Array(
1764                                 "img",
1765                                 "br",
1766                                 "hr",
1767                                 "input"
1768                                 );
1769
1770     $force_tag_closing = true;
1771
1772     $rm_attnames = Array(
1773                          "/.*/" =>
1774                          Array(
1775                                "/target/i",
1776                                "/^on.*/i",
1777                                "/^dynsrc/i",
1778                                "/^data.*/i",
1779                                "/^lowsrc.*/i"
1780                                )
1781                          );
1782
1783     $secremoveimg = "../images/" . _("sec_remove_eng.png");
1784     $bad_attvals = Array(
1785         "/.*/" =>
1786             Array(
1787                 "/^src|background/i" =>
1788                     Array(
1789                           Array(
1790                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1791                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1792                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1793                                 ),
1794                           Array(
1795                                 "\\1$secremoveimg\\2",
1796                                 "\\1$secremoveimg\\2",
1797                                 "\\1$secremoveimg\\2",
1798                                 "\\1$secremoveimg\\2"
1799                                 )
1800                         ),
1801                 "/^href|action/i" =>
1802                     Array(
1803                           Array(
1804                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1805                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1806                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1807                                 ),
1808                           Array(
1809                                 "\\1#\\1",
1810                                 "\\1#\\1",
1811                                 "\\1#\\1",
1812                                 "\\1#\\1"
1813                                 )
1814                         ),
1815                 "/^style/i" =>
1816                     Array(
1817                           Array(
1818                                 "/expression/i",
1819                                 "/binding/i",
1820                                 "/behaviou*r/i",
1821                                 "/include-source/i",
1822                                 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
1823                                 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
1824                                 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si"
1825                                ),
1826                           Array(
1827                                 "idiocy",
1828                                 "idiocy",
1829                                 "idiocy",
1830                                 "idiocy",
1831                                 "url(\\1#\\1)",
1832                                 "url(\\1#\\1)",
1833                                 "url(\\1#\\1)",
1834                                 "url(\\1#\\1)"
1835                                )
1836                           )
1837                 )
1838         );
1839     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1840         $view_unsafe_images = false;
1841     }
1842     if (!$view_unsafe_images){
1843         /**
1844          * Remove any references to http/https if view_unsafe_images set
1845          * to false.
1846          */
1847          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1848                     '/^([\'\"])\s*https*:.*([\'\"])/si');
1849          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1850                     "\\1$secremoveimg\\1");
1851          array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
1852                     '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1853          array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
1854                     "url(\\1$secremoveimg\\1)");
1855     }
1856
1857     $add_attr_to_tag = Array(
1858         "/^a$/i" =>
1859             Array('target'=>'"_new"',
1860                   'title'=>'"'._("This external link will open in a new window").'"'
1861             )
1862     );
1863     $trusted = sq_sanitize($body,
1864                            $tag_list,
1865                            $rm_tags_with_content,
1866                            $self_closing_tags,
1867                            $force_tag_closing,
1868                            $rm_attnames,
1869                            $bad_attvals,
1870                            $add_attr_to_tag,
1871                            $message,
1872                            $id,
1873                            $mailbox
1874                            );
1875     if (preg_match("|$secremoveimg|i", $trusted)){
1876         $has_unsafe_images = true;
1877     }
1878     return $trusted;
1879 }
1880
1881 /**
1882  * function SendDownloadHeaders - send file to the browser
1883  *
1884  * Original Source: SM core src/download.php
1885  * moved here to make it available to other code, and separate
1886  * front end from back end functionality.
1887  *
1888  * @param string $type0 first half of mime type
1889  * @param string $type1 second half of mime type
1890  * @param string $filename filename to tell the browser for downloaded file
1891  * @param boolean $force whether to force the download dialog to pop
1892  * @return void
1893  */
1894  function SendDownloadHeaders($type0, $type1, $filename, $force) {
1895      global $languages, $squirrelmail_language;
1896      $isIE = $isIE6 = 0;
1897
1898      sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
1899
1900      if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
1901          strstr($HTTP_USER_AGENT, 'Opera') === false) {
1902          $isIE = 1;
1903      }
1904
1905      if (strstr($HTTP_USER_AGENT, 'compatible; MSIE 6') !== false &&
1906          strstr($HTTP_USER_AGENT, 'Opera') === false) {
1907          $isIE6 = 1;
1908      }
1909
1910      if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
1911          function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
1912          $filename =
1913              $languages[$squirrelmail_language]['XTRA_CODE']('downloadfilename', $filename, $HTTP_USER_AGENT);
1914      } else {
1915         $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
1916      }
1917
1918      // A Pox on Microsoft and it's Office!
1919      if (!$force) {
1920          // Try to show in browser window
1921          header("Content-Disposition: inline; filename=\"$filename\"");
1922          header("Content-Type: $type0/$type1; name=\"$filename\"");
1923      } else {
1924          // Try to pop up the "save as" box
1925          // IE makes this hard.  It pops up 2 save boxes, or none.
1926          // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
1927          // But, accordint to Microsoft, it is "RFC compliant but doesn't
1928          // take into account some deviations that allowed within the
1929          // specification."  Doesn't that mean RFC non-compliant?
1930          // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
1931          //
1932          // The best thing you can do for IE is to upgrade to the latest
1933          // version
1934          if ($isIE && !$isIE6) {
1935              // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
1936              // Do not have quotes around filename, but that applied to
1937              // "attachment"... does it apply to inline too?
1938              //
1939              // This combination seems to work mostly.  IE 5.5 SP 1 has
1940              // known issues (see the Microsoft Knowledge Base)
1941              header("Content-Disposition: inline; filename=$filename");
1942              // This works for most types, but doesn't work with Word files
1943              header("Content-Type: application/download; name=\"$filename\"");
1944
1945              // These are spares, just in case.  :-)
1946              //header("Content-Type: $type0/$type1; name=\"$filename\"");
1947              //header("Content-Type: application/x-msdownload; name=\"$filename\"");
1948              //header("Content-Type: application/octet-stream; name=\"$filename\"");
1949          } else {
1950              header("Content-Disposition: attachment; filename=\"$filename\"");
1951              // application/octet-stream forces download for Netscape
1952              header("Content-Type: application/octet-stream; name=\"$filename\"");
1953          }
1954      }
1955  }
1956
1957 ?>