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 $uid_support;
 163
 164     /* Don't kill the connection if the browser is over a dialup
 165      * and it would take over 30 seconds to download it.
 166      * Don´t call set_time_limit in safe mode.
 167      */
 168
 169     if (!ini_get('safe_mode')) {
 170         set_time_limit(0);
 171     }
 172     /* in case of base64 encoded attachments, do not buffer them.
 173        Instead, echo the decoded attachment directly to screen */
 174     if (strtolower($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 = $i-1;
 617 //  if ($chunk{0} === '=') { /* performance, saves an unnessecarry preg call */
 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) {
 661             if ($htmlsave) {
 662                 $ret .= '&nbsp;';
 663             } else {
 664                 $ret .= ' ';
 665             }
 666         }
 667
 668         if (!$encoded && $htmlsave) {
 669             $ret .= htmlspecialchars($chunk);
 670         } else {
 671             $ret .= $chunk;
 672         }
 673         ++$i;
 674     }
 675     /* remove the first added space */
 676     if ($ret) {
 677         if ($htmlsave) {
 678             $ret = substr($ret,6);
 679         } else {
 680             $ret = substr($ret,1);
 681         }
 682     }
 683
 684     return $ret;
 685 }
 686
 687 /*
 688  * Encode a string according to RFC 1522 for use in headers if it
 689  * contains 8-bit characters or anything that looks like it should
 690  * be encoded.
 691  */
 692 function encodeHeader ($string) {
 693     global $default_charset, $languages, $squirrelmail_language;
 694
 695     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 696         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 697         return  $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
 698     }
 699     if (strtolower($default_charset) == 'iso-8859-1') {
 700         $string = str_replace("\240",' ',$string);
 701     }
 702
 703     // Encode only if the string contains 8-bit characters or =?
 704     $j = strlen($string);
 705     $max_l = 75 - strlen($default_charset) - 7;
 706     $aRet = array();
 707     $ret = '';
 708     $iEncStart = $enc_init = false;
 709     $cur_l = $iOffset = 0;
 710     for($i = 0; $i < $j; ++$i) {
 711         switch($string{$i})
 712         {
 713         case '=':
 714         case '<':
 715         case '>':
 716         case ',':
 717         case '?':
 718         case '_':
 719             if ($iEncStart === false) {
 720                 $iEncStart = $i;
 721             }
 722             $cur_l+=3;
 723             if ($cur_l > ($max_l-2)) {
 724                 /* if there is an stringpart that doesn't need encoding, add it */
 725                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 726                 $aRet[] = "=?$default_charset?Q?$ret?=";
 727                 $iOffset = $i;
 728                 $cur_l = 0;
 729                 $ret = '';
 730                 $iEncStart = false;
 731             } else {
 732                 $ret .= sprintf("=%02X",ord($string{$i}));
 733             }
 734             break;
 735         case '(':
 736         case ')':
 737             if ($iEncStart !== false) {
 738                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 739                 $aRet[] = "=?$default_charset?Q?$ret?=";
 740                 $iOffset = $i;
 741                 $cur_l = 0;
 742                 $ret = '';
 743                 $iEncStart = false;
 744             }
 745             break;
 746         case ' ':
 747             if ($iEncStart !== false) {
 748                 $cur_l++;
 749                 if ($cur_l > $max_l) {
 750                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 751                     $aRet[] = "=?$default_charset?Q?$ret?=";
 752                     $iOffset = $i;
 753                     $cur_l = 0;
 754                     $ret = '';
 755                     $iEncStart = false;
 756                 } else {
 757                     $ret .= '_';
 758                 }
 759             }
 760             break;
 761         default:
 762             $k = ord($string{$i});
 763             if ($k > 126) {
 764                 if ($iEncStart === false) {
 765                     // do not start encoding in the middle of a string, also take the rest of the word.
 766                     $sLeadString = substr($string,0,$i);
 767                     $aLeadString = explode(' ',$sLeadString);
 768                     $sToBeEncoded = array_pop($aLeadString);
 769                     $iEncStart = $i - strlen($sToBeEncoded);
 770                     $ret .= $sToBeEncoded;
 771                     $cur_l += strlen($sToBeEncoded);
 772                 }
 773                 $cur_l += 3;
 774                 /* first we add the encoded string that reached it's max size */
 775                 if ($cur_l > ($max_l-2)) {
 776                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 777                     $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
 778                     $cur_l = 3;
 779                     $ret = '';
 780                     $iOffset = $i;
 781                     $iEncStart = $i;
 782                 }
 783                 $enc_init = true;
 784                 $ret .= sprintf("=%02X", $k);
 785             } else {
 786                 if ($iEncStart !== false) {
 787                     $cur_l++;
 788                     if ($cur_l > $max_l) {
 789                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 790                         $aRet[] = "=?$default_charset?Q?$ret?=";
 791                         $iEncStart = false;
 792                         $iOffset = $i;
 793                         $cur_l = 0;
 794                         $ret = '';
 795                     } else {
 796                         $ret .= $string{$i};
 797                     }
 798                 }
 799             }
 800             break;
 801         }
 802     }
 803
 804     if ($enc_init) {
 805         if ($iEncStart !== false) {
 806             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 807             $aRet[] = "=?$default_charset?Q?$ret?=";
 808         } else {
 809             $aRet[] = substr($string,$iOffset);
 810         }
 811         $string = implode('',$aRet);
 812     }
 813     return $string;
 814 }
 815
 816 /* This function trys to locate the entity_id of a specific mime element */
 817 function find_ent_id($id, $message) {
 818     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
 819         if ($message->entities[$i]->header->type0 == 'multipart')  {
 820             $ret = find_ent_id($id, $message->entities[$i]);
 821         } else {
 822             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
 823 //                if (sq_check_save_extension($message->entities[$i])) {
 824                     return $message->entities[$i]->entity_id;
 825 //                }
 826             }
 827         }
 828     }
 829     return $ret;
 830 }
 831
 832 function sq_check_save_extension($message) {
 833     $filename = $message->getFilename();
 834     $ext = substr($filename, strrpos($filename,'.')+1);
 835     $save_extensions = array('jpg','jpeg','gif','png','bmp');
 836     return in_array($ext, $save_extensions);
 837 }
 838
 839
 840 /**
 841  ** HTMLFILTER ROUTINES
 842  */
 843
 844 /**
 845  * This function is more or less a wrapper around stripslashes. Apparently
 846  * Explorer is stupid enough to just remove the backslashes and then
 847  * execute the content of the attribute as if nothing happened.
 848  * Who does that?
 849  *
 850  * @param  attvalue   The value of the attribute
 851  * @return attvalue   The value of the attribute stripslashed.
 852  */
 853 function sq_unbackslash($attvalue){
 854     /**
 855      * Remove any backslashes. See if there are any first.
 856      */
 857     if (strstr($attvalue, '\\') !== false){
 858         $attvalue = stripslashes($attvalue);
 859     }
 860     return $attvalue;
 861 }
 862
 863 /**
 864  * Kill any tabs, newlines, or carriage returns. Our friends the
 865  * makers of the browser with 95% market value decided that it'd
 866  * be funny to make "java[tab]script" be just as good as "javascript".
 867  *
 868  * @param  attvalue  The attribute value before extraneous spaces removed.
 869  * @return attvalue  The attribute value after extraneous spaces removed.
 870  */
 871 function sq_unspace($attvalue){
 872     if (strcspn($attvalue, "\t\r\n") != strlen($attvalue)){
 873         $attvalue = str_replace(Array("\t", "\r", "\n"), Array('', '', ''),
 874                                 $attvalue);
 875     }
 876     return $attvalue;
 877 }
 878
 879 /**
 880  * This function returns the final tag out of the tag name, an array
 881  * of attributes, and the type of the tag. This function is called by
 882  * sq_sanitize internally.
 883  *
 884  * @param  $tagname  the name of the tag.
 885  * @param  $attary   the array of attributes and their values
 886  * @param  $tagtype  The type of the tag (see in comments).
 887  * @return           a string with the final tag representation.
 888  */
 889 function sq_tagprint($tagname, $attary, $tagtype){
 890     $me = 'sq_tagprint';
 891
 892     if ($tagtype == 2){
 893         $fulltag = '</' . $tagname . '>';
 894     } else {
 895         $fulltag = '<' . $tagname;
 896         if (is_array($attary) && sizeof($attary)){
 897             $atts = Array();
 898             while (list($attname, $attvalue) = each($attary)){
 899                 array_push($atts, "$attname=$attvalue");
 900             }
 901             $fulltag .= ' ' . join(" ", $atts);
 902         }
 903         if ($tagtype == 3){
 904             $fulltag .= ' /';
 905         }
 906         $fulltag .= '>';
 907     }
 908     return $fulltag;
 909 }
 910
 911 /**
 912  * A small helper function to use with array_walk. Modifies a by-ref
 913  * value and makes it lowercase.
 914  *
 915  * @param  $val a value passed by-ref.
 916  * @return      void since it modifies a by-ref value.
 917  */
 918 function sq_casenormalize(&$val){
 919     $val = strtolower($val);
 920 }
 921
 922 /**
 923  * This function skips any whitespace from the current position within
 924  * a string and to the next non-whitespace value.
 925  *
 926  * @param  $body   the string
 927  * @param  $offset the offset within the string where we should start
 928  *                 looking for the next non-whitespace character.
 929  * @return         the location within the $body where the next
 930  *                 non-whitespace char is located.
 931  */
 932 function sq_skipspace($body, $offset){
 933     $me = 'sq_skipspace';
 934     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
 935     if (sizeof($matches{1})){
 936         $count = strlen($matches{1});
 937         $offset += $count;
 938     }
 939     return $offset;
 940 }
 941
 942 /**
 943  * This function looks for the next character within a string.  It's
 944  * really just a glorified "strpos", except it catches if failures
 945  * nicely.
 946  *
 947  * @param  $body   The string to look for needle in.
 948  * @param  $offset Start looking from this position.
 949  * @param  $needle The character/string to look for.
 950  * @return         location of the next occurance of the needle, or
 951  *                 strlen($body) if needle wasn't found.
 952  */
 953 function sq_findnxstr($body, $offset, $needle){
 954     $me  = 'sq_findnxstr';
 955     $pos = strpos($body, $needle, $offset);
 956     if ($pos === FALSE){
 957         $pos = strlen($body);
 958     }
 959     return $pos;
 960 }
 961
 962 /**
 963  * This function takes a PCRE-style regexp and tries to match it
 964  * within the string.
 965  *
 966  * @param  $body   The string to look for needle in.
 967  * @param  $offset Start looking from here.
 968  * @param  $reg    A PCRE-style regex to match.
 969  * @return         Returns a false if no matches found, or an array
 970  *                 with the following members:
 971  *                 - integer with the location of the match within $body
 972  *                 - string with whatever content between offset and the match
 973  *                 - string with whatever it is we matched
 974  */
 975 function sq_findnxreg($body, $offset, $reg){
 976     $me = 'sq_findnxreg';
 977     $matches = Array();
 978     $retarr = Array();
 979     preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
 980     if (!isset($matches{0}) || !$matches{0}){
 981         $retarr = false;
 982     } else {
 983         $retarr{0} = $offset + strlen($matches{1});
 984         $retarr{1} = $matches{1};
 985         $retarr{2} = $matches{2};
 986     }
 987     return $retarr;
 988 }
 989
 990 /**
 991  * This function looks for the next tag.
 992  *
 993  * @param  $body   String where to look for the next tag.
 994  * @param  $offset Start looking from here.
 995  * @return         false if no more tags exist in the body, or
 996  *                 an array with the following members:
 997  *                 - string with the name of the tag
 998  *                 - array with attributes and their values
 999  *                 - integer with tag type (1, 2, or 3)
1000  *                 - integer where the tag starts (starting "<")
1001  *                 - integer where the tag ends (ending ">")
1002  *                 first three members will be false, if the tag is invalid.
1003  */
1004 function sq_getnxtag($body, $offset){
1005     $me = 'sq_getnxtag';
1006     if ($offset > strlen($body)){
1007         return false;
1008     }
1009     $lt = sq_findnxstr($body, $offset, "<");
1010     if ($lt == strlen($body)){
1011         return false;
1012     }
1013     /**
1014      * We are here:
1015      * blah blah <tag attribute="value">
1016      * \---------^
1017      */
1018     $pos = sq_skipspace($body, $lt+1);
1019     if ($pos >= strlen($body)){
1020         return Array(false, false, false, $lt, strlen($body));
1021     }
1022     /**
1023      * There are 3 kinds of tags:
1024      * 1. Opening tag, e.g.:
1025      *    <a href="blah">
1026      * 2. Closing tag, e.g.:
1027      *    </a>
1028      * 3. XHTML-style content-less tag, e.g.:
1029      *    <img src="blah"/>
1030      */
1031     $tagtype = false;
1032     switch (substr($body, $pos, 1)){
1033         case '/':
1034             $tagtype = 2;
1035             $pos++;
1036             break;
1037         case '!':
1038             /**
1039              * A comment or an SGML declaration.
1040              */
1041             if (substr($body, $pos+1, 2) == "--"){
1042                 $gt = strpos($body, "-->", $pos);
1043                 if ($gt === false){
1044                     $gt = strlen($body);
1045                 } else {
1046                     $gt += 2;
1047                 }
1048                 return Array(false, false, false, $lt, $gt);
1049             } else {
1050                 $gt = sq_findnxstr($body, $pos, ">");
1051                 return Array(false, false, false, $lt, $gt);
1052             }
1053             break;
1054         default:
1055             /**
1056              * Assume tagtype 1 for now. If it's type 3, we'll switch values
1057              * later.
1058              */
1059             $tagtype = 1;
1060             break;
1061     }
1062
1063     $tag_start = $pos;
1064     $tagname = '';
1065     /**
1066      * Look for next [\W-_], which will indicate the end of the tag name.
1067      */
1068     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1069     if ($regary == false){
1070         return Array(false, false, false, $lt, strlen($body));
1071     }
1072     list($pos, $tagname, $match) = $regary;
1073     $tagname = strtolower($tagname);
1074
1075     /**
1076      * $match can be either of these:
1077      * '>'  indicating the end of the tag entirely.
1078      * '\s' indicating the end of the tag name.
1079      * '/'  indicating that this is type-3 xhtml tag.
1080      *
1081      * Whatever else we find there indicates an invalid tag.
1082      */
1083     switch ($match){
1084         case '/':
1085             /**
1086              * This is an xhtml-style tag with a closing / at the
1087              * end, like so: <img src="blah"/>. Check if it's followed
1088              * by the closing bracket. If not, then this tag is invalid
1089              */
1090             if (substr($body, $pos, 2) == "/>"){
1091                 $pos++;
1092                 $tagtype = 3;
1093             } else {
1094                 $gt = sq_findnxstr($body, $pos, ">");
1095                 $retary = Array(false, false, false, $lt, $gt);
1096                 return $retary;
1097             }
1098         case '>':
1099             return Array($tagname, false, $tagtype, $lt, $pos);
1100             break;
1101         default:
1102             /**
1103              * Check if it's whitespace
1104              */
1105             if (!preg_match('/\s/', $match)){
1106                 /**
1107                  * This is an invalid tag! Look for the next closing ">".
1108                  */
1109                 $gt = sq_findnxstr($body, $lt, ">");
1110                 return Array(false, false, false, $lt, $gt);
1111             }
1112             break;
1113     }
1114
1115     /**
1116      * At this point we're here:
1117      * <tagname  attribute='blah'>
1118      * \-------^
1119      *
1120      * At this point we loop in order to find all attributes.
1121      */
1122     $attname = '';
1123     $atttype = false;
1124     $attary = Array();
1125
1126     while ($pos <= strlen($body)){
1127         $pos = sq_skipspace($body, $pos);
1128         if ($pos == strlen($body)){
1129             /**
1130              * Non-closed tag.
1131              */
1132             return Array(false, false, false, $lt, $pos);
1133         }
1134         /**
1135          * See if we arrived at a ">" or "/>", which means that we reached
1136          * the end of the tag.
1137          */
1138         $matches = Array();
1139         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1140             /**
1141              * Yep. So we did.
1142              */
1143             $pos += strlen($matches{1});
1144             if ($matches{2} == "/>"){
1145                 $tagtype = 3;
1146                 $pos++;
1147             }
1148             return Array($tagname, $attary, $tagtype, $lt, $pos);
1149         }
1150
1151         /**
1152          * There are several types of attributes, with optional
1153          * [:space:] between members.
1154          * Type 1:
1155          *   attrname[:space:]=[:space:]'CDATA'
1156          * Type 2:
1157          *   attrname[:space:]=[:space:]"CDATA"
1158          * Type 3:
1159          *   attr[:space:]=[:space:]CDATA
1160          * Type 4:
1161          *   attrname
1162          *
1163          * We leave types 1 and 2 the same, type 3 we check for
1164          * '"' and convert to "&quot" if needed, then wrap in
1165          * double quotes. Type 4 we convert into:
1166          * attrname="yes".
1167          */
1168         $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1169         if ($regary == false){
1170             /**
1171              * Looks like body ended before the end of tag.
1172              */
1173             return Array(false, false, false, $lt, strlen($body));
1174         }
1175         list($pos, $attname, $match) = $regary;
1176         $attname = strtolower($attname);
1177         /**
1178          * We arrived at the end of attribute name. Several things possible
1179          * here:
1180          * '>'  means the end of the tag and this is attribute type 4
1181          * '/'  if followed by '>' means the same thing as above
1182          * '\s' means a lot of things -- look what it's followed by.
1183          *      anything else means the attribute is invalid.
1184          */
1185         switch($match){
1186             case '/':
1187                 /**
1188                  * This is an xhtml-style tag with a closing / at the
1189                  * end, like so: <img src="blah"/>. Check if it's followed
1190                  * by the closing bracket. If not, then this tag is invalid
1191                  */
1192                 if (substr($body, $pos, 2) == "/>"){
1193                     $pos++;
1194                     $tagtype = 3;
1195                 } else {
1196                     $gt = sq_findnxstr($body, $pos, ">");
1197                     $retary = Array(false, false, false, $lt, $gt);
1198                     return $retary;
1199                 }
1200             case '>':
1201                 $attary{$attname} = '"yes"';
1202                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1203                 break;
1204             default:
1205                 /**
1206                  * Skip whitespace and see what we arrive at.
1207                  */
1208                 $pos = sq_skipspace($body, $pos);
1209                 $char = substr($body, $pos, 1);
1210                 /**
1211                  * Two things are valid here:
1212                  * '=' means this is attribute type 1 2 or 3.
1213                  * \w means this was attribute type 4.
1214                  * anything else we ignore and re-loop. End of tag and
1215                  * invalid stuff will be caught by our checks at the beginning
1216                  * of the loop.
1217                  */
1218                 if ($char == "="){
1219                     $pos++;
1220                     $pos = sq_skipspace($body, $pos);
1221                     /**
1222                      * Here are 3 possibilities:
1223                      * "'"  attribute type 1
1224                      * '"'  attribute type 2
1225                      * everything else is the content of tag type 3
1226                      */
1227                     $quot = substr($body, $pos, 1);
1228                     if ($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} = "'" . $attval . "'";
1236                     } else if ($quot == '"'){
1237                         $regary = sq_findnxreg($body, $pos+1, '\"');
1238                         if ($regary == false){
1239                             return Array(false, false, false, $lt, strlen($body));
1240                         }
1241                         list($pos, $attval, $match) = $regary;
1242                         $pos++;
1243                         $attary{$attname} = '"' . $attval . '"';
1244                     } else {
1245                         /**
1246                          * These are hateful. Look for \s, or >.
1247                          */
1248                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1249                         if ($regary == false){
1250                             return Array(false, false, false, $lt, strlen($body));
1251                         }
1252                         list($pos, $attval, $match) = $regary;
1253                         /**
1254                          * If it's ">" it will be caught at the top.
1255                          */
1256                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1257                         $attary{$attname} = '"' . $attval . '"';
1258                     }
1259                 } else if (preg_match("|[\w/>]|", $char)) {
1260                     /**
1261                      * That was attribute type 4.
1262                      */
1263                     $attary{$attname} = '"yes"';
1264                 } else {
1265                     /**
1266                      * An illegal character. Find next '>' and return.
1267                      */
1268                     $gt = sq_findnxstr($body, $pos, ">");
1269                     return Array(false, false, false, $lt, $gt);
1270                 }
1271                 break;
1272         }
1273     }
1274     /**
1275      * The fact that we got here indicates that the tag end was never
1276      * found. Return invalid tag indication so it gets stripped.
1277      */
1278     return Array(false, false, false, $lt, strlen($body));
1279 }
1280
1281 /**
1282  * This function checks attribute values for entity-encoded values
1283  * and returns them translated into 8-bit strings so we can run
1284  * checks on them.
1285  *
1286  * @param  $attvalue A string to run entity check against.
1287  * @return           Translated value.
1288  */
1289 function sq_deent($attvalue){
1290     $me = 'sq_deent';
1291     /**
1292      * See if we have to run the checks first. All entities must start
1293      * with "&".
1294      */
1295     if (strpos($attvalue, "&") === false){
1296         return $attvalue;
1297     }
1298     /**
1299      * Check named entities first.
1300      */
1301     $trans = get_html_translation_table(HTML_ENTITIES);
1302     /**
1303      * Leave &quot; in, as it can mess us up.
1304      */
1305     $trans = array_flip($trans);
1306     unset($trans{"&quot;"});
1307     while (list($ent, $val) = each($trans)){
1308         $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1309     }
1310     /**
1311      * Now translate numbered entities from 1 to 255 if needed.
1312      */
1313     if (strpos($attvalue, "#") !== false){
1314         $omit = Array(34, 39);
1315         for ($asc=1; $asc<256; $asc++){
1316             if (!in_array($asc, $omit)){
1317                 $chr = chr($asc);
1318                 $attvalue = preg_replace("/\&#0*$asc;*(\D)/si", "$chr\\1",
1319                                          $attvalue);
1320                 $attvalue = preg_replace("/\&#x0*".dechex($asc).";*(\W)/si",
1321                                          "$chr\\1", $attvalue);
1322             }
1323         }
1324     }
1325     return $attvalue;
1326 }
1327
1328 /**
1329  * This function runs various checks against the attributes.
1330  *
1331  * @param  $tagname         String with the name of the tag.
1332  * @param  $attary          Array with all tag attributes.
1333  * @param  $rm_attnames     See description for sq_sanitize
1334  * @param  $bad_attvals     See description for sq_sanitize
1335  * @param  $add_attr_to_tag See description for sq_sanitize
1336  * @param  $message         message object
1337  * @param  $id              message id
1338  * @return                  Array with modified attributes.
1339  */
1340 function sq_fixatts($tagname,
1341                     $attary,
1342                     $rm_attnames,
1343                     $bad_attvals,
1344                     $add_attr_to_tag,
1345                     $message,
1346                     $id,
1347                     $mailbox
1348                     ){
1349     $me = 'sq_fixatts';
1350     while (list($attname, $attvalue) = each($attary)){
1351         /**
1352          * See if this attribute should be removed.
1353          */
1354         foreach ($rm_attnames as $matchtag=>$matchattrs){
1355             if (preg_match($matchtag, $tagname)){
1356                 foreach ($matchattrs as $matchattr){
1357                     if (preg_match($matchattr, $attname)){
1358                         unset($attary{$attname});
1359                         continue;
1360                     }
1361                 }
1362             }
1363         }
1364         /**
1365          * Remove any backslashes, entities, and extraneous whitespace.
1366          */
1367         $attvalue = sq_unbackslash($attvalue);
1368         $attvalue = sq_deent($attvalue);
1369         $attvalue = sq_unspace($attvalue);
1370
1371         /**
1372          * Now let's run checks on the attvalues.
1373          * I don't expect anyone to comprehend this. If you do,
1374          * get in touch with me so I can drive to where you live and
1375          * shake your hand personally. :)
1376          */
1377         foreach ($bad_attvals as $matchtag=>$matchattrs){
1378             if (preg_match($matchtag, $tagname)){
1379                 foreach ($matchattrs as $matchattr=>$valary){
1380                     if (preg_match($matchattr, $attname)){
1381                         /**
1382                          * There are two arrays in valary.
1383                          * First is matches.
1384                          * Second one is replacements
1385                          */
1386                         list($valmatch, $valrepl) = $valary;
1387                         $newvalue =
1388                             preg_replace($valmatch, $valrepl, $attvalue);
1389                         if ($newvalue != $attvalue){
1390                             $attary{$attname} = $newvalue;
1391                         }
1392                     }
1393                 }
1394             }
1395         }
1396         /**
1397          * Turn cid: urls into http-friendly ones.
1398          */
1399         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1400             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1401         }
1402     }
1403     /**
1404      * See if we need to append any attributes to this tag.
1405      */
1406     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1407         if (preg_match($matchtag, $tagname)){
1408             $attary = array_merge($attary, $addattary);
1409         }
1410     }
1411     return $attary;
1412 }
1413
1414 /**
1415  * This function edits the style definition to make them friendly and
1416  * usable in squirrelmail.
1417  *
1418  * @param  $message  the message object
1419  * @param  $id       the message id
1420  * @param  $content  a string with whatever is between <style> and </style>
1421  * @return           a string with edited content.
1422  */
1423 function sq_fixstyle($body, $pos, $message, $id){
1424     global $view_unsafe_images;
1425     $me = 'sq_fixstyle';
1426     $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1427     if ($ret == FALSE){
1428         return array(FALSE, strlen($body));
1429     }
1430     $newpos = $ret[0] + strlen($ret[2]);
1431     $content = $ret[1];
1432     /**
1433      * First look for general BODY style declaration, which would be
1434      * like so:
1435      * body {background: blah-blah}
1436      * and change it to .bodyclass so we can just assign it to a <div>
1437      */
1438     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1439     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1440     /**
1441      * Fix url('blah') declarations.
1442      */
1443     $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1444                             "url(\\1$secremoveimg\\2)", $content);
1445     /**
1446      * Fix url('https*://.*) declarations but only if $view_unsafe_images
1447      * is false.
1448      */
1449     if (!$view_unsafe_images){
1450         $content = preg_replace("|url\s*\(\s*([\'\"])\s*https*:.*?([\'\"])\s*\)|si",
1451                                 "url(\\1$secremoveimg\\2)", $content);
1452     }
1453
1454     /**
1455      * Fix urls that refer to cid:
1456      */
1457     while (preg_match("|url\s*\(\s*([\'\"]\s*cid:.*?[\'\"])\s*\)|si",
1458                       $content, $matches)){
1459         $cidurl = $matches{1};
1460         $httpurl = sq_cid2http($message, $id, $cidurl);
1461         $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1462                                 "url($httpurl)", $content);
1463     }
1464
1465     /**
1466      * Fix stupid css declarations which lead to vulnerabilities
1467      * in IE.
1468      */
1469     $match   = Array('/expression/i',
1470                      '/behaviou*r/i',
1471                      '/binding/i',
1472                      '/include-source/i');
1473     $replace = Array('idiocy', 'idiocy', 'idiocy', 'idiocy');
1474     $content = preg_replace($match, $replace, $content);
1475     return array($content, $newpos);
1476 }
1477
1478 /**
1479  * This function converts cid: url's into the ones that can be viewed in
1480  * the browser.
1481  *
1482  * @param  $message  the message object
1483  * @param  $id       the message id
1484  * @param  $cidurl   the cid: url.
1485  * @return           a string with a http-friendly url
1486  */
1487 function sq_cid2http($message, $id, $cidurl, $mailbox){
1488     /**
1489      * Get rid of quotes.
1490      */
1491     $quotchar = substr($cidurl, 0, 1);
1492     if ($quotchar == '"' || $quotchar == "'"){
1493         $cidurl = str_replace($quotchar, "", $cidurl);
1494     } else {
1495         $quotchar = '';
1496     }
1497     $cidurl = substr(trim($cidurl), 4);
1498     $linkurl = find_ent_id($cidurl, $message);
1499     /* in case of non-save cid links $httpurl should be replaced by a sort of
1500        unsave link image */
1501     $httpurl = '';
1502     if ($linkurl) {
1503         $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1504                    "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1505                    '&amp;ent_id=' . $linkurl . $quotchar;
1506     }
1507     return $httpurl;
1508 }
1509
1510 /**
1511  * This function changes the <body> tag into a <div> tag since we
1512  * can't really have a body-within-body.
1513  *
1514  * @param  $attary   an array of attributes and values of <body>
1515  * @param  $mailbox  mailbox we're currently reading (for cid2http)
1516  * @param  $message  current message (for cid2http)
1517  * @param  $id       current message id (for cid2http)
1518  * @return           a modified array of attributes to be set for <div>
1519  */
1520 function sq_body2div($attary, $mailbox, $message, $id){
1521     $me = 'sq_body2div';
1522     $divattary = Array('class' => "'bodyclass'");
1523     $bgcolor = '#ffffff';
1524     $text = '#000000';
1525     $has_bgc_stl = $has_txt_stl = false;
1526     $styledef = '';
1527     if (is_array($attary) && sizeof($attary) > 0){
1528         foreach ($attary as $attname=>$attvalue){
1529             $quotchar = substr($attvalue, 0, 1);
1530             $attvalue = str_replace($quotchar, "", $attvalue);
1531             switch ($attname){
1532                 case 'background':
1533                     $attvalue = sq_cid2http($message, $id,
1534                                             $attvalue, $mailbox);
1535                     $styledef .= "background-image: url('$attvalue'); ";
1536                     break;
1537                 case 'bgcolor':
1538                     $has_bgc_stl = true;
1539                     $styledef .= "background-color: $attvalue; ";
1540                     break;
1541                 case 'text':
1542                     $has_txt_stl = true;
1543                     $styledef .= "color: $attvalue; ";
1544                     break;
1545             }
1546         }
1547         // Outlook defines a white bgcolor and no text color. This can lead to
1548         // white text on a white bg with certain themes.
1549         if ($has_bgc_stl && !$has_txt_stl) {
1550             $styledef .= "color: $text; ";
1551         }
1552         if (strlen($styledef) > 0){
1553             $divattary{"style"} = "\"$styledef\"";
1554         }
1555     }
1556     return $divattary;
1557 }
1558
1559 /**
1560  * This is the main function and the one you should actually be calling.
1561  * There are several variables you should be aware of an which need
1562  * special description.
1563  *
1564  * Since the description is quite lengthy, see it here:
1565  * http://www.mricon.com/html/phpfilter.html
1566  *
1567  * @param $body                 the string with HTML you wish to filter
1568  * @param $tag_list             see description above
1569  * @param $rm_tags_with_content see description above
1570  * @param $self_closing_tags    see description above
1571  * @param $force_tag_closing    see description above
1572  * @param $rm_attnames          see description above
1573  * @param $bad_attvals          see description above
1574  * @param $add_attr_to_tag      see description above
1575  * @param $message              message object
1576  * @param $id                   message id
1577  * @return                      sanitized html safe to show on your pages.
1578  */
1579 function sq_sanitize($body,
1580                      $tag_list,
1581                      $rm_tags_with_content,
1582                      $self_closing_tags,
1583                      $force_tag_closing,
1584                      $rm_attnames,
1585                      $bad_attvals,
1586                      $add_attr_to_tag,
1587                      $message,
1588                      $id,
1589                      $mailbox
1590                      ){
1591     $me = 'sq_sanitize';
1592     $rm_tags = array_shift($tag_list);
1593     /**
1594      * Normalize rm_tags and rm_tags_with_content.
1595      */
1596     @array_walk($tag_list, 'sq_casenormalize');
1597     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1598     @array_walk($self_closing_tags, 'sq_casenormalize');
1599     /**
1600      * See if tag_list is of tags to remove or tags to allow.
1601      * false  means remove these tags
1602      * true   means allow these tags
1603      */
1604     $curpos = 0;
1605     $open_tags = Array();
1606     $trusted = "\n<!-- begin sanitized html -->\n";
1607     $skip_content = false;
1608     /**
1609      * Take care of netscape's stupid javascript entities like
1610      * &{alert('boo')};
1611      */
1612     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1613
1614     while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1615         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1616         $free_content = substr($body, $curpos, $lt-$curpos);
1617         /**
1618          * Take care of <style>
1619          */
1620         if ($tagname == "style" && $tagtype == 1){
1621             list($free_content, $curpos) =
1622                 sq_fixstyle($body, $gt+1, $message, $id);
1623             if ($free_content != FALSE){
1624                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1625                 $trusted .= $free_content;
1626                 $trusted .= sq_tagprint($tagname, false, 2);
1627             }
1628             continue;
1629         }
1630         if ($skip_content == false){
1631             $trusted .= $free_content;
1632         }
1633         if ($tagname != FALSE){
1634             if ($tagtype == 2){
1635                 if ($skip_content == $tagname){
1636                     /**
1637                      * Got to the end of tag we needed to remove.
1638                      */
1639                     $tagname = false;
1640                     $skip_content = false;
1641                 } else {
1642                     if ($skip_content == false){
1643                         if ($tagname == "body"){
1644                             $tagname = "div";
1645                         }
1646                         if (isset($open_tags{$tagname}) &&
1647                             $open_tags{$tagname} > 0){
1648                             $open_tags{$tagname}--;
1649                         } else {
1650                             $tagname = false;
1651                         }
1652                     }
1653                 }
1654             } else {
1655                 /**
1656                  * $rm_tags_with_content
1657                  */
1658                 if ($skip_content == false){
1659                     /**
1660                      * See if this is a self-closing type and change
1661                      * tagtype appropriately.
1662                      */
1663                     if ($tagtype == 1
1664                         && in_array($tagname, $self_closing_tags)){
1665                         $tagtype = 3;
1666                     }
1667                     /**
1668                      * See if we should skip this tag and any content
1669                      * inside it.
1670                      */
1671                     if ($tagtype == 1 &&
1672                         in_array($tagname, $rm_tags_with_content)){
1673                         $skip_content = $tagname;
1674                     } else {
1675                         if (($rm_tags == false
1676                              && in_array($tagname, $tag_list)) ||
1677                             ($rm_tags == true &&
1678                              !in_array($tagname, $tag_list))){
1679                             $tagname = false;
1680                         } else {
1681                             /**
1682                              * Convert body into div.
1683                              */
1684                             if ($tagname == "body"){
1685                                 $tagname = "div";
1686                                 $attary = sq_body2div($attary, $mailbox,
1687                                                       $message, $id);
1688                             }
1689                             if ($tagtype == 1){
1690                                 if (isset($open_tags{$tagname})){
1691                                     $open_tags{$tagname}++;
1692                                 } else {
1693                                     $open_tags{$tagname}=1;
1694                                 }
1695                             }
1696                             /**
1697                              * This is where we run other checks.
1698                              */
1699                             if (is_array($attary) && sizeof($attary) > 0){
1700                                 $attary = sq_fixatts($tagname,
1701                                                      $attary,
1702                                                      $rm_attnames,
1703                                                      $bad_attvals,
1704                                                      $add_attr_to_tag,
1705                                                      $message,
1706                                                      $id,
1707                                                      $mailbox
1708                                                      );
1709                             }
1710                         }
1711                     }
1712                 }
1713             }
1714             if ($tagname != false && $skip_content == false){
1715                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1716             }
1717         }
1718         $curpos = $gt+1;
1719     }
1720     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1721     if ($force_tag_closing == true){
1722         foreach ($open_tags as $tagname=>$opentimes){
1723             while ($opentimes > 0){
1724                 $trusted .= '</' . $tagname . '>';
1725                 $opentimes--;
1726             }
1727         }
1728         $trusted .= "\n";
1729     }
1730     $trusted .= "<!-- end sanitized html -->\n";
1731     return $trusted;
1732 }
1733
1734 /**
1735  * This is a wrapper function to call html sanitizing routines.
1736  *
1737  * @param  $body  the body of the message
1738  * @param  $id    the id of the message
1739  * @return        a string with html safe to display in the browser.
1740  */
1741 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1742     global $attachment_common_show_images, $view_unsafe_images,
1743            $has_unsafe_images;
1744     /**
1745      * Don't display attached images in HTML mode.
1746      */
1747     $attachment_common_show_images = false;
1748     $tag_list = Array(
1749                       false,
1750                       "object",
1751                       "meta",
1752                       "html",
1753                       "head",
1754                       "base",
1755                       "link",
1756                       "frame",
1757                       "iframe",
1758                       "plaintext",
1759                       "marquee"
1760                       );
1761
1762     $rm_tags_with_content = Array(
1763                                   "script",
1764                                   "applet",
1765                                   "embed",
1766                                   "title",
1767                                   "frameset",
1768                                   "xml"
1769                                   );
1770
1771     $self_closing_tags =  Array(
1772                                 "img",
1773                                 "br",
1774                                 "hr",
1775                                 "input"
1776                                 );
1777
1778     $force_tag_closing = true;
1779
1780     $rm_attnames = Array(
1781                          "/.*/" =>
1782                          Array(
1783                                "/target/i",
1784                                "/^on.*/i",
1785                                "/^dynsrc/i",
1786                                "/^data.*/i",
1787                                "/^lowsrc.*/i"
1788                                )
1789                          );
1790
1791     $secremoveimg = "../images/" . _("sec_remove_eng.png");
1792     $bad_attvals = Array(
1793         "/.*/" =>
1794             Array(
1795                 "/^src|background/i" =>
1796                     Array(
1797                           Array(
1798                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1799                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1800                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1801                                 ),
1802                           Array(
1803                                 "\\1$secremoveimg\\2",
1804                                 "\\1$secremoveimg\\2",
1805                                 "\\1$secremoveimg\\2",
1806                                 "\\1$secremoveimg\\2"
1807                                 )
1808                         ),
1809                 "/^href|action/i" =>
1810                     Array(
1811                           Array(
1812                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1813                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1814                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1815                                 ),
1816                           Array(
1817                                 "\\1#\\1",
1818                                 "\\1#\\1",
1819                                 "\\1#\\1",
1820                                 "\\1#\\1"
1821                                 )
1822                         ),
1823                 "/^style/i" =>
1824                     Array(
1825                           Array(
1826                                 "/expression/i",
1827                                 "/binding/i",
1828                                 "/behaviou*r/i",
1829                                 "/include-source/i",
1830                                 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
1831                                 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
1832                                 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si"
1833                                ),
1834                           Array(
1835                                 "idiocy",
1836                                 "idiocy",
1837                                 "idiocy",
1838                                 "idiocy",
1839                                 "url(\\1#\\1)",
1840                                 "url(\\1#\\1)",
1841                                 "url(\\1#\\1)",
1842                                 "url(\\1#\\1)"
1843                                )
1844                           )
1845                 )
1846         );
1847     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1848         $view_unsafe_images = false;
1849     }
1850     if (!$view_unsafe_images){
1851         /**
1852          * Remove any references to http/https if view_unsafe_images set
1853          * to false.
1854          */
1855          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1856                     '/^([\'\"])\s*https*:.*([\'\"])/si');
1857          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1858                     "\\1$secremoveimg\\1");
1859          array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
1860                     '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1861          array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
1862                     "url(\\1$secremoveimg\\1)");
1863     }
1864
1865     $add_attr_to_tag = Array(
1866         "/^a$/i" =>
1867             Array('target'=>'"_new"',
1868                   'title'=>'"'._("This external link will open in a new window").'"'
1869             )
1870     );
1871     $trusted = sq_sanitize($body,
1872                            $tag_list,
1873                            $rm_tags_with_content,
1874                            $self_closing_tags,
1875                            $force_tag_closing,
1876                            $rm_attnames,
1877                            $bad_attvals,
1878                            $add_attr_to_tag,
1879                            $message,
1880                            $id,
1881                            $mailbox
1882                            );
1883     if (preg_match("|$secremoveimg|i", $trusted)){
1884         $has_unsafe_images = true;
1885     }
1886     return $trusted;
1887 }
1888
1889 /**
1890  * function SendDownloadHeaders - send file to the browser
1891  *
1892  * Original Source: SM core src/download.php
1893  * moved here to make it available to other code, and separate
1894  * front end from back end functionality.
1895  *
1896  * @param string $type0 first half of mime type
1897  * @param string $type1 second half of mime type
1898  * @param string $filename filename to tell the browser for downloaded file
1899  * @param boolean $force whether to force the download dialog to pop
1900  * @return void
1901  */
1902  function SendDownloadHeaders($type0, $type1, $filename, $force) {
1903      global $languages, $squirrelmail_language;
1904      $isIE = $isIE6 = 0;
1905
1906      sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
1907
1908      if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
1909          strstr($HTTP_USER_AGENT, 'Opera') === false) {
1910          $isIE = 1;
1911      }
1912
1913      if (strstr($HTTP_USER_AGENT, 'compatible; MSIE 6') !== false &&
1914          strstr($HTTP_USER_AGENT, 'Opera') === false) {
1915          $isIE6 = 1;
1916      }
1917
1918      if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
1919          function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
1920          $filename =
1921              $languages[$squirrelmail_language]['XTRA_CODE']('downloadfilename', $filename, $HTTP_USER_AGENT);
1922      } else {
1923         $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace('&nbsp;', ' ', $filename));
1924      }
1925
1926      // A Pox on Microsoft and it's Office!
1927      if (!$force) {
1928          // Try to show in browser window
1929          header("Content-Disposition: inline; filename=\"$filename\"");
1930          header("Content-Type: $type0/$type1; name=\"$filename\"");
1931      } else {
1932          // Try to pop up the "save as" box
1933          // IE makes this hard.  It pops up 2 save boxes, or none.
1934          // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
1935          // But, accordint to Microsoft, it is "RFC compliant but doesn't
1936          // take into account some deviations that allowed within the
1937          // specification."  Doesn't that mean RFC non-compliant?
1938          // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
1939          //
1940          // The best thing you can do for IE is to upgrade to the latest
1941          // version
1942          if ($isIE && !$isIE6) {
1943              // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
1944              // Do not have quotes around filename, but that applied to
1945              // "attachment"... does it apply to inline too?
1946              //
1947              // This combination seems to work mostly.  IE 5.5 SP 1 has
1948              // known issues (see the Microsoft Knowledge Base)
1949              header("Content-Disposition: inline; filename=$filename");
1950              // This works for most types, but doesn't work with Word files
1951              header("Content-Type: application/download; name=\"$filename\"");
1952
1953              // These are spares, just in case.  :-)
1954              //header("Content-Type: $type0/$type1; name=\"$filename\"");
1955              //header("Content-Type: application/x-msdownload; name=\"$filename\"");
1956              //header("Content-Type: application/octet-stream; name=\"$filename\"");
1957          } else {
1958              header("Content-Disposition: attachment; filename=\"$filename\"");
1959              // application/octet-stream forces download for Netscape
1960              header("Content-Type: application/octet-stream; name=\"$filename\"");
1961          }
1962      }
1963  }
1964
1965 ?>