functions/mime.php

   1 <?php
   2
   3 /**
   4  * mime.php
   5  *
   6  * Copyright (c) 1999-2002 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  */
  14
  15 require_once('../functions/imap.php');
  16 require_once('../functions/attachment_common.php');
  17
  18 /* --------------------------------------------------------------------------------- */
  19 /* MIME DECODING                                                                     */
  20 /* --------------------------------------------------------------------------------- */
  21
  22 /* This function gets the structure of a message and stores it in the "message" class.
  23  * It will return this object for use with all relevant header information and
  24  * fully parsed into the standard "message" object format.
  25  */
  26
  27 function mime_structure ($bodystructure, $flags=array()) {
  28
  29     // isolate the body structure and remove beginning and end parenthesis
  30     $read = trim(substr ($bodystructure, strpos(strtolower($bodystructure), 'bodystructure') + 13));
  31     $msg =& new Message();
  32     $read = trim(substr ($read, 0, -1));
  33     $res = $msg->parseStructure($read);
  34     $msg = $res[0];
  35     if (!is_object($msg)) {
  36         include_once( '../functions/display_messages.php' );
  37         global $color, $mailbox;
  38         displayPageHeader( $color, urldecode($mailbox) );
  39         echo "<BODY TEXT=\"$color[8]\" BGCOLOR=\"$color[4]\" LINK=\"$color[7]\" VLINK=\"$color[7]\" ALINK=\"$color[7]\">\n\n" .
  40          '<CENTER>';
  41         $errormessage = _("Squirrelmail could not decode the bodystructure of the message");
  42         $errormessage .= '<BR>'._("the provided bodystructure by your imap-server").':<BR><BR>';
  43         $errormessage .= '<table><tr><td>'.htmlspecialchars($read).'</td></tr></table>';
  44         plain_error_message( $errormessage, $color );
  45         echo '</body></html>';
  46         exit;
  47     }
  48     $msg->setEnt('0');
  49     if (count($flags)) {
  50         foreach ($flags as $flag) {
  51             $char = strtoupper($flag{1});
  52             switch ($char) {
  53             case 'S':
  54                 if (strtolower($flag) == '\\seen') {
  55                     $msg->is_seen = true;
  56                 }
  57                 break;
  58             case 'A':
  59                 if (strtolower($flag) == '\\answered') {
  60                     $msg->is_answered = true;
  61                 }
  62                 break;
  63             case 'D':
  64                 if (strtolower($flag) == '\\deleted') {
  65                     $msg->is_deleted = true;
  66                 }
  67                 break;
  68             case 'F':
  69                 if (strtolower($flag) == '\\flagged') {
  70                     $msg->is_flagged = true;
  71                 }
  72                 break;
  73             case 'M':
  74                 if (strtolower($flag) == '$mdnsent') {
  75                     $msg->is_mdnsent = true;
  76                 }
  77                 break;
  78             default:
  79                 break;
  80             }
  81         }
  82     }
  83     //    listEntities($msg);
  84     return( $msg );
  85 }
  86
  87 /* this starts the parsing of a particular structure.  It is called recursively,
  88  * so it can be passed different structures.  It returns an object of type
  89  * $message.
  90  * First, it checks to see if it is a multipart message.  If it is, then it
  91  * handles that as it sees is necessary.  If it is just a regular entity,
  92  * then it parses it and adds the necessary header information (by calling out
  93  * to mime_get_elements()
  94  */
  95
  96 function mime_fetch_body($imap_stream, $id, $ent_id) {
  97     global $uid_support;
  98     /*
  99      * do a bit of error correction.  If we couldn't find the entity id, just guess
 100      * that it is the first one.  That is usually the case anyway.
 101      */
 102     if (!$ent_id) {
 103         $ent_id = 1;
 104     }
 105     $cmd = "FETCH $id BODY[$ent_id]";
 106
 107     $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, $uid_support);
 108     do {
 109         $topline = trim(array_shift( $data ));
 110     } while( $topline && $topline[0] == '*' && !preg_match( '/\* [0-9]+ FETCH.*/i', $topline )) ;
 111
 112     $wholemessage = implode('', $data);
 113     if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
 114
 115         $ret = substr( $wholemessage, 0, $regs[1] );
 116         /*
 117             There is some information in the content info header that could be important
 118             in order to parse html messages. Let's get them here.
 119         */
 120         if ( $ret{0} == '<' ) {
 121             $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, $uid_support);
 122         }
 123     } else if (ereg('"([^"]*)"', $topline, $regs)) {
 124         $ret = $regs[1];
 125     } else {
 126         global $where, $what, $mailbox, $passed_id, $startMessage;
 127         $par = 'mailbox=' . urlencode($mailbox) . "&amp;passed_id=$passed_id";
 128         if (isset($where) && isset($what)) {
 129             $par .= '&amp;where='. urlencode($where) . "&amp;what=" . urlencode($what);
 130         } else {
 131             $par .= "&amp;startMessage=$startMessage&amp;show_more=0";
 132         }
 133         $par .= '&amp;response=' . urlencode($response) .
 134                 '&amp;message=' . urlencode($message).
 135                 '&amp;topline=' . urlencode($topline);
 136
 137         echo   '<tt><br>' .
 138                '<table width="80%"><tr>' .
 139                '<tr><td colspan=2>' .
 140                _("Body retrieval error. The reason for this is most probably that the message is malformed. Please help us making future versions better by submitting this message to the developers knowledgebase!") .
 141                " <A HREF=\"../src/retrievalerror.php?$par\"><br>" .
 142                _("Submit message") . '</A><BR>&nbsp;' .
 143                '</td></tr>' .
 144                '<td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
 145                '<td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
 146                '<td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
 147                '<td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
 148                "</table><BR></tt></font><hr>";
 149
 150         $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, $uid_support);
 151         array_shift($data);
 152         $wholemessage = implode('', $data);
 153
 154         $ret = $wholemessage;
 155     }
 156     return( $ret );
 157 }
 158
 159 function mime_print_body_lines ($imap_stream, $id, $ent_id, $encoding) {
 160     global $uid_support;
 161     // do a bit of error correction.  If we couldn't find the entity id, just guess
 162     // that it is the first one.  That is usually the case anyway.
 163     if (!$ent_id) {
 164         $ent_id = 1;
 165     }
 166     $sid = sqimap_session_id($uid_support);
 167     // Don't kill the connection if the browser is over a dialup
 168     // and it would take over 30 seconds to download it.
 169
 170     // don´t call set_time_limit in safe mode.
 171     if (!ini_get("safe_mode")) {
 172         set_time_limit(0);
 173     }
 174     if ($uid_support) {
 175        $sid_s = substr($sid,0,strpos($sid, ' '));
 176     } else {
 177        $sid_s = $sid;
 178     }
 179
 180     $body = mime_fetch_body ($imap_stream, $id, $ent_id);
 181     echo decodeBody($body, $encoding);
 182     return;
 183 /*
 184     fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
 185     $cnt = 0;
 186     $continue = true;
 187     $read = fgets ($imap_stream,8192);
 188
 189
 190     // This could be bad -- if the section has sqimap_session_id() . ' OK'
 191     // or similar, it will kill the download.
 192     while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
 193         if (trim($read) == ')==') {
 194             $read1 = $read;
 195             $read = fgets ($imap_stream,4096);
 196             if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
 197                 return;
 198             } else {
 199                 echo decodeBody($read1, $encoding) .
 200                      decodeBody($read, $encoding);
 201             }
 202         } else if ($cnt) {
 203             echo decodeBody($read, $encoding);
 204         }
 205         $read = fgets ($imap_stream,4096);
 206         $cnt++;
 207 //      break;
 208     }
 209 */
 210 }
 211
 212 /* -[ END MIME DECODING ]----------------------------------------------------------- */
 213
 214 // This is here for debugging purposese.  It will print out a list
 215 // of all the entity IDs that are in the $message object.
 216
 217 function listEntities ($message) {
 218   if ($message) {
 219     echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br>';
 220     for ($i = 0;isset($message->entities[$i]); $i++) {
 221        echo "$i : ";
 222        $msg = listEntities($message->entities[$i]);
 223
 224        if ($msg) {
 225           echo "return: ";
 226           return $msg;
 227        }
 228     }
 229   }
 230 }
 231
 232 function getPriorityStr($priority) {
 233    $priority_level = substr($priority,0,1);
 234
 235    switch($priority_level) {
 236      /* check for a higher then normal priority. */
 237      case '1':
 238      case '2':
 239         $priority_string = _("High");
 240         break;
 241
 242       /* check for a lower then normal priority. */
 243      case '4':
 244      case '5':
 245         $priority_string = _("Low");
 246         break;
 247
 248      /* check for a normal priority. */
 249      case '3':
 250      default:
 251         $priority_level = '3';
 252         $priority_string = _("Normal");
 253         break;
 254
 255    }
 256    return $priority_string;
 257 }
 258
 259 /* returns a $message object for a particular entity id */
 260 function getEntity ($message, $ent_id) {
 261     return $message->getEntity($ent_id);
 262 }
 263
 264 /*
 265  * translateText
 266  * Extracted from strings.php 23/03/2002
 267  */
 268
 269 function translateText(&$body, $wrap_at, $charset) {
 270     global $where, $what; /* from searching */
 271     global $color; /* color theme */
 272
 273     require_once('../functions/url_parser.php');
 274
 275     $body_ary = explode("\n", $body);
 276     $PriorQuotes = 0;
 277     for ($i=0; $i < count($body_ary); $i++) {
 278         $line = $body_ary[$i];
 279         if (strlen($line) - 2 >= $wrap_at) {
 280             sqWordWrap($line, $wrap_at);
 281         }
 282         $line = charset_decode($charset, $line);
 283         $line = str_replace("\t", '        ', $line);
 284
 285         parseUrl ($line);
 286
 287         $Quotes = 0;
 288         $pos = 0;
 289         $j = strlen( $line );
 290
 291         while ( $pos < $j ) {
 292             if ($line[$pos] == ' ') {
 293                 $pos ++;
 294             } else if (strpos($line, '&gt;', $pos) === $pos) {
 295                 $pos += 4;
 296                 $Quotes ++;
 297             } else {
 298                 break;
 299             }
 300         }
 301
 302         if ($Quotes > 1) {
 303             if (! isset($color[14])) {
 304                 $color[14] = '#FF0000';
 305             }
 306             $line = '<FONT COLOR="' . $color[14] . '">' . $line . '</FONT>';
 307         } elseif ($Quotes) {
 308             if (! isset($color[13])) {
 309                 $color[13] = '#800000';
 310             }
 311             $line = '<FONT COLOR="' . $color[13] . '">' . $line . '</FONT>';
 312         }
 313
 314         $body_ary[$i] = $line;
 315     }
 316     $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
 317 }
 318
 319
 320 /* This returns a parsed string called $body. That string can then
 321 be displayed as the actual message in the HTML. It contains
 322 everything needed, including HTML Tags, Attachments at the
 323 bottom, etc.
 324 */
 325 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
 326     // this if statement checks for the entity to show as the
 327     // primary message. To add more of them, just put them in the
 328     // order that is their priority.
 329     global $startMessage, $username, $key, $imapServerAddress, $imapPort,
 330            $show_html_default, $has_unsafe_images, $view_unsafe_images, $sort;
 331
 332     $has_unsafe_images= 0;
 333     $body = '';
 334     $urlmailbox = urlencode($mailbox);
 335     $body_message = getEntity($message, $ent_num);
 336     if (($body_message->header->type0 == 'text') ||
 337         ($body_message->header->type0 == 'rfc822')) {
 338             $body = mime_fetch_body ($imap_stream, $id, $ent_num);
 339         $body = decodeBody($body, $body_message->header->encoding);
 340         $hookResults = do_hook("message_body", $body);
 341         $body = $hookResults[1];
 342
 343         // If there are other types that shouldn't be formatted, add
 344         // them here
 345
 346         if ($body_message->header->type1 == 'html') {
 347             if ( $show_html_default <> 1 ) {
 348                 $entity_conv = array('&nbsp;' => ' ',
 349                                      '&gt;'   => '>',
 350                                      '&lt;'   => '<');
 351                 $body = strip_tags( $body );
 352                 $body = strtr($body, $entity_conv);
 353                 $body = trim($body);
 354                 translateText($body, $wrap_at,
 355                               $body_message->header->getParameter('charset'));
 356             } else {
 357                 $body = magicHTML( $body, $id, $message, $mailbox );
 358             }
 359         } else {
 360             translateText($body, $wrap_at,
 361                               $body_message->header->getParameter('charset'));
 362         }
 363
 364         if ($has_unsafe_images) {
 365             if ($view_unsafe_images) {
 366                 $body .= "<CENTER><SMALL><A HREF=\"read_body.php?passed_id=$id&amp;passed_ent_id=".$message->entity_id."&amp;mailbox=$urlmailbox&amp;sort=$sort&amp;startMessage=$startMessage&amp;show_more=0\">". _("Hide Unsafe Images") ."</A></SMALL></CENTER><BR>\n";
 367             } else {
 368                 $body .= "<CENTER><SMALL><A HREF=\"read_body.php?passed_id=$id&amp;passed_ent_id=".$message->entity_id."&amp;mailbox=$urlmailbox&amp;sort=$sort&amp;startMessage=$startMessage&amp;show_more=0&amp;view_unsafe_images=1\">". _("View Unsafe Images") ."</A></SMALL></CENTER><BR>\n";
 369             }
 370         }
 371     }
 372     return ($body);
 373 }
 374
 375
 376 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 377     global $where, $what, $startMessage, $color;
 378     static $ShownHTML = 0;
 379
 380     $att_ar = $message->getAttachments($exclude_id);
 381
 382     if (!count($att_ar)) return '';
 383
 384     $attachments = '';
 385
 386     $urlMailbox = urlencode($mailbox);
 387
 388     foreach ($att_ar as $att) {
 389         $ent = urldecode($att->entity_id);
 390         $header = $att->header;
 391         $type0 = strtolower($header->type0);
 392         $type1 = strtolower($header->type1);
 393         $name = '';
 394         $Links['download link']['text'] = _("download");
 395         $Links['download link']['href'] =
 396                 "../src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 397         $ImageURL = '';
 398         if ($type0 =='message' && $type1 == 'rfc822') {
 399             $default_page = '../src/read_body.php';
 400             $rfc822_header = $att->rfc822_header;
 401             $filename = decodeHeader($rfc822_header->subject);
 402
 403             $from_o = $rfc822_header->from;
 404             if (is_object($from_o)) {
 405                 $from_name = $from_o->getAddress(false);
 406             } else {
 407                 $from_name = _("Unknown sender");
 408             }
 409             $from_name = decodeHeader(htmlspecialchars($from_name));
 410             $description = $from_name;
 411         } else {
 412             $default_page = '../src/download.php';
 413             if (is_object($header->disposition)) {
 414                $filename = decodeHeader($header->disposition->getProperty('filename'));
 415                if (trim($filename) == '') {
 416                   $name = decodeHeader($header->disposition->getProperty('name'));
 417                   if (trim($name) == '') {
 418                      if ( trim( $header->id ) == '' )
 419                         $filename = 'untitled-[' . $ent . ']' ;
 420                      else
 421                         $filename = 'cid: ' . $header->id;
 422                   } else {
 423                      $filename = $name;
 424                   }
 425                }
 426             } else {
 427                if ( trim( $header->id ) == '' )
 428                   $filename = 'untitled-[' . $ent . ']' ;
 429                else
 430                   $filename = 'cid: ' . $header->id;
 431             }
 432
 433             if ($header->description) {
 434                 $description = htmlspecialchars($header->description);
 435             } else {
 436                 $description = '';
 437             }
 438         }
 439
 440         $display_filename = $filename;
 441         if (isset($passed_ent_id)) {
 442            $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 443         } else {
 444            $passed_ent_id_link = '';
 445         }
 446         $DefaultLink = $default_page . "?startMessage=$startMessage"
 447                      . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 448                      . '&amp;ent_id='.$ent.$passed_ent_id_link;
 449         if ($where && $what) {
 450            $DefaultLink = '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 451         }
 452         /* this executes the attachment hook with a specific MIME-type.
 453             * if that doens't have results, it tries if there's a rule
 454             * for a more generic type. */
 455         $HookResults = do_hook("attachment $type0/$type1", $Links,
 456             $startMessage, $id, $urlMailbox, $ent, $DefaultLink,
 457             $display_filename, $where, $what);
 458         if(count($HookResults[1]) <= 1) {
 459             $HookResults = do_hook("attachment $type0/*", $Links,
 460                     $startMessage, $id, $urlMailbox, $ent, $DefaultLink,
 461                     $display_filename, $where, $what);
 462         }
 463
 464         $Links = $HookResults[1];
 465         $DefaultLink = $HookResults[6];
 466
 467         $attachments .= '<TR><TD>' .
 468                         "<A HREF=\"$DefaultLink\">$display_filename</A>&nbsp;</TD>" .
 469                         '<TD><SMALL><b>' . show_readable_size($header->size) .
 470                         '</b>&nbsp;&nbsp;</small></TD>' .
 471                         "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
 472                         '<TD><SMALL>';
 473         $attachments .= '<b>' . $description . '</b>';
 474         $attachments .= '</SMALL></TD><TD><SMALL>&nbsp;';
 475
 476         $SkipSpaces = 1;
 477         foreach ($Links as $Val) {
 478             if ($SkipSpaces) {
 479                 $SkipSpaces = 0;
 480             } else {
 481                 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
 482             }
 483             $attachments .= '<a href="' . $Val['href'] . '">' .  $Val['text'] . '</a>';
 484         }
 485         unset($Links);
 486         $attachments .= "</TD></TR>\n";
 487     }
 488     return $attachments;
 489 }
 490
 491 /** this function decodes the body depending on the encoding type. **/
 492 function decodeBody($body, $encoding) {
 493     global $languages, $squirrelmail_language;
 494
 495     $body = str_replace("\r\n", "\n", $body);
 496     $encoding = strtolower($encoding);
 497
 498     global $show_html_default;
 499
 500     if ($encoding == 'quoted-printable' ||
 501     $encoding == 'quoted_printable') {
 502         $body = quoted_printable_decode($body);
 503
 504         while (ereg("=\n", $body)) {
 505             $body = ereg_replace ("=\n", '', $body);
 506         }
 507
 508     } else if ($encoding == 'base64') {
 509         $body = base64_decode($body);
 510     }
 511
 512     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 513         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 514         $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
 515     }
 516
 517     // All other encodings are returned raw.
 518     return( $body );
 519 }
 520
 521 /*
 522  * This functions decode strings that is encoded according to
 523  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 524  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 525  */
 526 function decodeHeader ($string, $utfencode=true) {
 527     global $languages, $squirrelmail_language;
 528     if (is_array($string)) {
 529         $string = implode("\n", $string);
 530     }
 531
 532     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 533         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 534         $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
 535     }
 536
 537     $i = 0;
 538     while (preg_match('/^(.{' . $i . '})(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=/Ui',
 539                       $string, $res)) {
 540         $prefix = $res[1];
 541         // Ignore white-space between consecutive encoded-words
 542         if (strspn($res[2], " \t") != strlen($res[2])) {
 543             $prefix .= $res[2];
 544         }
 545
 546         if (ucfirst($res[4]) == 'B') {
 547             $replace = base64_decode($res[5]);
 548         } else {
 549             $replace = str_replace('_', ' ', $res[5]);
 550             $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 551                                     $replace);
 552             /* Only encode into entities by default. Some places
 553                don't need the encoding, like the compose form. */
 554             if ($utfencode) {
 555                 $replace = charset_decode($res[3], $replace);
 556             }
 557         }
 558         $string = $prefix . $replace . substr($string, strlen($res[0]));
 559         $i = strlen($prefix) + strlen($replace);
 560     }
 561     return( $string );
 562 }
 563
 564 /*
 565  * Encode a string according to RFC 1522 for use in headers if it
 566  * contains 8-bit characters or anything that looks like it should
 567  * be encoded.
 568  */
 569 function encodeHeader ($string) {
 570     global $default_charset, $languages, $squirrelmail_language;
 571
 572     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 573         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 574         return  $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
 575     }
 576
 577     // Encode only if the string contains 8-bit characters or =?
 578     $j = strlen( $string  );
 579     $l = strstr($string, '=?');         // Must be encoded ?
 580     $ret = '';
 581     for( $i=0; $i < $j; ++$i) {
 582         switch( $string{$i} ) {
 583            case '=':
 584           $ret .= '=3D';
 585           break;
 586         case '?':
 587           $ret .= '=3F';
 588           break;
 589         case '_':
 590           $ret .= '=5F';
 591           break;
 592         case ' ':
 593           $ret .= '_';
 594           break;
 595         default:
 596           $k = ord( $string{$i} );
 597           if ( $k > 126 ) {
 598              $ret .= sprintf("=%02X", $k);
 599              $l = TRUE;
 600           } else
 601              $ret .= $string{$i};
 602         }
 603     }
 604
 605     if ( $l ) {
 606         $string = "=?$default_charset?Q?$ret?=";
 607     }
 608
 609     return( $string );
 610 }
 611
 612 /* This function trys to locate the entity_id of a specific mime element */
 613
 614 function find_ent_id( $id, $message ) {
 615     $ret = '';
 616     for ($i=0; $ret == '' && $i < count($message->entities); $i++) {
 617         if ( $message->entities[$i]->header->type0 == 'multipart')  {
 618             $ret = find_ent_id( $id, $message->entities[$i] );
 619         } else {
 620             if ( strcasecmp( $message->entities[$i]->header->id, $id ) == 0 ) {
 621                 if (sq_check_save_extension($message->entities[$i])) {
 622                     $ret = $message->entities[$i]->entity_id;
 623                 } else {
 624                     $ret = '';
 625                 }
 626             }
 627         }
 628     }
 629     return( $ret );
 630 }
 631
 632 function sq_check_save_extension($message) {
 633     $filename = $message->getFilename();
 634     $ext = substr($filename, strrpos($filename,'.')+1);
 635     $save_extensions = array('jpg','jpeg','gif','png','bmp');
 636     return (in_array($ext, $save_extensions));
 637 }
 638
 639
 640 /**
 641  ** HTMLFILTER ROUTINES
 642  */
 643
 644 /**
 645  * This function returns the final tag out of the tag name, an array
 646  * of attributes, and the type of the tag. This function is called by
 647  * sq_sanitize internally.
 648  *
 649  * @param  $tagname  the name of the tag.
 650  * @param  $attary   the array of attributes and their values
 651  * @param  $tagtype  The type of the tag (see in comments).
 652  * @return           a string with the final tag representation.
 653  */
 654 function sq_tagprint($tagname, $attary, $tagtype){
 655
 656     $me = 'sq_tagprint';
 657
 658     if ($tagtype == 2){
 659         $fulltag = '</' . $tagname . '>';
 660     } else {
 661         $fulltag = '<' . $tagname;
 662         if (is_array($attary) && sizeof($attary)){
 663             $atts = Array();
 664             while (list($attname, $attvalue) = each($attary)){
 665                 array_push($atts, "$attname=$attvalue");
 666             }
 667             $fulltag .= ' ' . join(" ", $atts);
 668         }
 669         if ($tagtype == 3){
 670             $fulltag .= ' /';
 671         }
 672         $fulltag .= '>';
 673     }
 674     return $fulltag;
 675 }
 676
 677 /**
 678  * A small helper function to use with array_walk. Modifies a by-ref
 679  * value and makes it lowercase.
 680  *
 681  * @param  $val a value passed by-ref.
 682  * @return      void since it modifies a by-ref value.
 683  */
 684 function sq_casenormalize(&$val){
 685     $val = strtolower($val);
 686 }
 687
 688 /**
 689  * This function skips any whitespace from the current position within
 690  * a string and to the next non-whitespace value.
 691  *
 692  * @param  $body   the string
 693  * @param  $offset the offset within the string where we should start
 694  *                 looking for the next non-whitespace character.
 695  * @return         the location within the $body where the next
 696  *                 non-whitespace char is located.
 697  */
 698 function sq_skipspace($body, $offset){
 699     $me = 'sq_skipspace';
 700     preg_match("/^(\s*)/s", substr($body, $offset), $matches);
 701     if (sizeof($matches{1})){
 702         $count = strlen($matches{1});
 703         $offset += $count;
 704     }
 705     return $offset;
 706 }
 707
 708 /**
 709  * This function looks for the next character within a string.  It's
 710  * really just a glorified "strpos", except it catches if failures
 711  * nicely.
 712  *
 713  * @param  $body   The string to look for needle in.
 714  * @param  $offset Start looking from this position.
 715  * @param  $needle The character/string to look for.
 716  * @return         location of the next occurance of the needle, or
 717  *                 strlen($body) if needle wasn't found.
 718  */
 719 function sq_findnxstr($body, $offset, $needle){
 720     $me = 'sq_findnxstr';
 721     $pos = strpos($body, $needle, $offset);
 722     if ($pos === FALSE){
 723         $pos = strlen($body);
 724     }
 725     return $pos;
 726 }
 727
 728 /**
 729  * This function takes a PCRE-style regexp and tries to match it
 730  * within the string.
 731  *
 732  * @param  $body   The string to look for needle in.
 733  * @param  $offset Start looking from here.
 734  * @param  $reg    A PCRE-style regex to match.
 735  * @return         Returns a false if no matches found, or an array
 736  *                 with the following members:
 737  *                 - integer with the location of the match within $body
 738  *                 - string with whatever content between offset and the match
 739  *                 - string with whatever it is we matched
 740  */
 741 function sq_findnxreg($body, $offset, $reg){
 742     $me = 'sq_findnxreg';
 743     $matches = Array();
 744     $retarr = Array();
 745     preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
 746     if (!$matches{0}){
 747         $retarr = false;
 748     } else {
 749         $retarr{0} = $offset + strlen($matches{1});
 750         $retarr{1} = $matches{1};
 751         $retarr{2} = $matches{2};
 752     }
 753     return $retarr;
 754 }
 755
 756 /**
 757  * This function looks for the next tag.
 758  *
 759  * @param  $body   String where to look for the next tag.
 760  * @param  $offset Start looking from here.
 761  * @return         false if no more tags exist in the body, or
 762  *                 an array with the following members:
 763  *                 - string with the name of the tag
 764  *                 - array with attributes and their values
 765  *                 - integer with tag type (1, 2, or 3)
 766  *                 - integer where the tag starts (starting "<")
 767  *                 - integer where the tag ends (ending ">")
 768  *                 first three members will be false, if the tag is invalid.
 769  */
 770 function sq_getnxtag($body, $offset){
 771     $me = 'sq_getnxtag';
 772     if ($offset > strlen($body)){
 773         return false;
 774     }
 775     $lt = sq_findnxstr($body, $offset, "<");
 776     if ($lt == strlen($body)){
 777         return false;
 778     }
 779     /**
 780      * We are here:
 781      * blah blah <tag attribute="value">
 782      * \---------^
 783      */
 784     $pos = sq_skipspace($body, $lt+1);
 785     if ($pos >= strlen($body)){
 786         return Array(false, false, false, $lt, strlen($body));
 787     }
 788     /**
 789      * There are 3 kinds of tags:
 790      * 1. Opening tag, e.g.:
 791      *    <a href="blah">
 792      * 2. Closing tag, e.g.:
 793      *    </a>
 794      * 3. XHTML-style content-less tag, e.g.:
 795      *    <img src="blah"/>
 796      */
 797     $tagtype = false;
 798     switch (substr($body, $pos, 1)){
 799     case '/':
 800         $tagtype = 2;
 801         $pos++;
 802         break;
 803     case '!':
 804         /**
 805          * A comment or an SGML declaration.
 806          */
 807         if (substr($body, $pos+1, 2) == "--"){
 808             $gt = strpos($body, "-->", $pos);
 809             if ($gt === false){
 810                 $gt = strlen($body);
 811             } else {
 812                 $gt += 2;
 813             }
 814             return Array(false, false, false, $lt, $gt);
 815         } else {
 816             $gt = sq_findnxstr($body, $pos, ">");
 817             return Array(false, false, false, $lt, $gt);
 818         }
 819         break;
 820     default:
 821         /**
 822          * Assume tagtype 1 for now. If it's type 3, we'll switch values
 823          * later.
 824          */
 825         $tagtype = 1;
 826         break;
 827     }
 828
 829     $tag_start = $pos;
 830     $tagname = '';
 831     /**
 832      * Look for next [\W-_], which will indicate the end of the tag name.
 833      */
 834     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
 835     if ($regary == false){
 836         return Array(false, false, false, $lt, strlen($body));
 837     }
 838     list($pos, $tagname, $match) = $regary;
 839     $tagname = strtolower($tagname);
 840
 841     /**
 842      * $match can be either of these:
 843      * '>'  indicating the end of the tag entirely.
 844      * '\s' indicating the end of the tag name.
 845      * '/'  indicating that this is type-3 xhtml tag.
 846      *
 847      * Whatever else we find there indicates an invalid tag.
 848      */
 849     switch ($match){
 850     case '/':
 851         /**
 852          * This is an xhtml-style tag with a closing / at the
 853          * end, like so: <img src="blah"/>. Check if it's followed
 854          * by the closing bracket. If not, then this tag is invalid
 855          */
 856         if (substr($body, $pos, 2) == "/>"){
 857             $pos++;
 858             $tagtype = 3;
 859         } else {
 860             $gt = sq_findnxstr($body, $pos, ">");
 861             $retary = Array(false, false, false, $lt, $gt);
 862             return $retary;
 863         }
 864     case '>':
 865         return Array($tagname, false, $tagtype, $lt, $pos);
 866         break;
 867     default:
 868         /**
 869          * Check if it's whitespace
 870          */
 871         if (preg_match("/\s/", $match)){
 872         } else {
 873             /**
 874              * This is an invalid tag! Look for the next closing ">".
 875              */
 876             $gt = sq_findnxstr($body, $offset, ">");
 877             return Array(false, false, false, $lt, $gt);
 878         }
 879     }
 880
 881     /**
 882      * At this point we're here:
 883      * <tagname  attribute='blah'>
 884      * \-------^
 885      *
 886      * At this point we loop in order to find all attributes.
 887      */
 888     $attname = '';
 889     $atttype = false;
 890     $attary = Array();
 891
 892     while ($pos <= strlen($body)){
 893         $pos = sq_skipspace($body, $pos);
 894         if ($pos == strlen($body)){
 895             /**
 896              * Non-closed tag.
 897              */
 898             return Array(false, false, false, $lt, $pos);
 899         }
 900         /**
 901          * See if we arrived at a ">" or "/>", which means that we reached
 902          * the end of the tag.
 903          */
 904         $matches = Array();
 905         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
 906             /**
 907              * Yep. So we did.
 908              */
 909             $pos += strlen($matches{1});
 910             if ($matches{2} == "/>"){
 911                 $tagtype = 3;
 912                 $pos++;
 913             }
 914             return Array($tagname, $attary, $tagtype, $lt, $pos);
 915         }
 916
 917         /**
 918          * There are several types of attributes, with optional
 919          * [:space:] between members.
 920          * Type 1:
 921          *   attrname[:space:]=[:space:]'CDATA'
 922          * Type 2:
 923          *   attrname[:space:]=[:space:]"CDATA"
 924          * Type 3:
 925          *   attr[:space:]=[:space:]CDATA
 926          * Type 4:
 927          *   attrname
 928          *
 929          * We leave types 1 and 2 the same, type 3 we check for
 930          * '"' and convert to "&quot" if needed, then wrap in
 931          * double quotes. Type 4 we convert into:
 932          * attrname="yes".
 933          */
 934         $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
 935         if ($regary == false){
 936             /**
 937              * Looks like body ended before the end of tag.
 938              */
 939             return Array(false, false, false, $lt, strlen($body));
 940         }
 941         list($pos, $attname, $match) = $regary;
 942         $attname = strtolower($attname);
 943         /**
 944          * We arrived at the end of attribute name. Several things possible
 945          * here:
 946          * '>'  means the end of the tag and this is attribute type 4
 947          * '/'  if followed by '>' means the same thing as above
 948          * '\s' means a lot of things -- look what it's followed by.
 949          *      anything else means the attribute is invalid.
 950          */
 951         switch($match){
 952         case '/':
 953             /**
 954              * This is an xhtml-style tag with a closing / at the
 955              * end, like so: <img src="blah"/>. Check if it's followed
 956              * by the closing bracket. If not, then this tag is invalid
 957              */
 958             if (substr($body, $pos, 2) == "/>"){
 959                 $pos++;
 960                 $tagtype = 3;
 961             } else {
 962                 $gt = sq_findnxstr($body, $pos, ">");
 963                 $retary = Array(false, false, false, $lt, $gt);
 964                 return $retary;
 965             }
 966         case '>':
 967             $attary{$attname} = '"yes"';
 968             return Array($tagname, $attary, $tagtype, $lt, $pos);
 969             break;
 970         default:
 971             /**
 972              * Skip whitespace and see what we arrive at.
 973              */
 974             $pos = sq_skipspace($body, $pos);
 975             $char = substr($body, $pos, 1);
 976             /**
 977              * Two things are valid here:
 978              * '=' means this is attribute type 1 2 or 3.
 979              * \w means this was attribute type 4.
 980              * anything else we ignore and re-loop. End of tag and
 981              * invalid stuff will be caught by our checks at the beginning
 982              * of the loop.
 983              */
 984             if ($char == "="){
 985                 $pos++;
 986                 $pos = sq_skipspace($body, $pos);
 987                 /**
 988                  * Here are 3 possibilities:
 989                  * "'"  attribute type 1
 990                  * '"'  attribute type 2
 991                  * everything else is the content of tag type 3
 992                  */
 993                 $quot = substr($body, $pos, 1);
 994                 if ($quot == "'"){
 995                     $regary = sq_findnxreg($body, $pos+1, "\'");
 996                     if ($regary == false){
 997                         return Array(false, false, false, $lt, strlen($body));
 998                     }
 999                     list($pos, $attval, $match) = $regary;
1000                     $pos++;
1001                     $attary{$attname} = "'" . $attval . "'";
1002                 } else if ($quot == '"'){
1003                     $regary = sq_findnxreg($body, $pos+1, '\"');
1004                     if ($regary == false){
1005                         return Array(false, false, false, $lt, strlen($body));
1006                     }
1007                     list($pos, $attval, $match) = $regary;
1008                     $pos++;
1009                     $attary{$attname} = '"' . $attval . '"';
1010                 } else {
1011                     /**
1012                      * These are hateful. Look for \s, or >.
1013                      */
1014                     $regary = sq_findnxreg($body, $pos, "[\s>]");
1015                     if ($regary == false){
1016                         return Array(false, false, false, $lt, strlen($body));
1017                     }
1018                     list($pos, $attval, $match) = $regary;
1019                     /**
1020                      * If it's ">" it will be caught at the top.
1021                      */
1022                     $attval = preg_replace("/\"/s", "&quot;", $attval);
1023                     $attary{$attname} = '"' . $attval . '"';
1024                 }
1025             } else if (preg_match("|[\w/>]|", $char)) {
1026                 /**
1027                  * That was attribute type 4.
1028                  */
1029                 $attary{$attname} = '"yes"';
1030             } else {
1031                 /**
1032                  * An illegal character. Find next '>' and return.
1033                  */
1034                 $gt = sq_findnxstr($body, $pos, ">");
1035                 return Array(false, false, false, $lt, $gt);
1036             }
1037         }
1038     }
1039     /**
1040      * The fact that we got here indicates that the tag end was never
1041      * found. Return invalid tag indication so it gets stripped.
1042      */
1043     return Array(false, false, false, $lt, strlen($body));
1044 }
1045
1046 /**
1047  * This function checks attribute values for entity-encoded values
1048  * and returns them translated into 8-bit strings so we can run
1049  * checks on them.
1050  *
1051  * @param  $attvalue A string to run entity check against.
1052  * @return           Translated value.
1053  */
1054 function sq_deent($attvalue){
1055     $me = 'sq_deent';
1056     /**
1057      * See if we have to run the checks first. All entities must start
1058      * with "&".
1059      */
1060     if (strpos($attvalue, "&") === false){
1061         return $attvalue;
1062     }
1063     /**
1064      * Check named entities first.
1065      */
1066     $trans = get_html_translation_table(HTML_ENTITIES);
1067     /**
1068      * Leave &quot; in, as it can mess us up.
1069      */
1070     $trans = array_flip($trans);
1071     unset($trans{"&quot;"});
1072     while (list($ent, $val) = each($trans)){
1073         $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1074     }
1075     /**
1076      * Now translate numbered entities from 1 to 255 if needed.
1077      */
1078     if (strpos($attvalue, "#") !== false){
1079         $omit = Array(34, 39);
1080         for ($asc=1; $asc<256; $asc++){
1081             if (!in_array($asc, $omit)){
1082                 $chr = chr($asc);
1083                 $attvalue = preg_replace("/\&#0*$asc;*(\D)/si", "$chr\\1",
1084                                          $attvalue);
1085                 $attvalue = preg_replace("/\&#x0*".dechex($asc).";*(\W)/si",
1086                                          "$chr\\1", $attvalue);
1087             }
1088         }
1089     }
1090     return $attvalue;
1091 }
1092
1093 /**
1094  * This function runs various checks against the attributes.
1095  *
1096  * @param  $tagname         String with the name of the tag.
1097  * @param  $attary          Array with all tag attributes.
1098  * @param  $rm_attnames     See description for sq_sanitize
1099  * @param  $bad_attvals     See description for sq_sanitize
1100  * @param  $add_attr_to_tag See description for sq_sanitize
1101  * @param  $message         message object
1102  * @param  $id              message id
1103  * @return                  Array with modified attributes.
1104  */
1105 function sq_fixatts($tagname,
1106                     $attary,
1107                     $rm_attnames,
1108                     $bad_attvals,
1109                     $add_attr_to_tag,
1110                     $message,
1111                     $id,
1112                     $mailbox
1113                     ){
1114     $me = 'sq_fixatts';
1115     while (list($attname, $attvalue) = each($attary)){
1116         /**
1117          * See if this attribute should be removed.
1118          */
1119         foreach ($rm_attnames as $matchtag=>$matchattrs){
1120             if (preg_match($matchtag, $tagname)){
1121                 foreach ($matchattrs as $matchattr){
1122                     if (preg_match($matchattr, $attname)){
1123                         unset($attary{$attname});
1124                         continue;
1125                     }
1126                 }
1127             }
1128         }
1129         /**
1130          * Remove any entities.
1131          */
1132         $attvalue = sq_deent($attvalue);
1133
1134         /**
1135          * Now let's run checks on the attvalues.
1136          * I don't expect anyone to comprehend this. If you do,
1137          * get in touch with me so I can drive to where you live and
1138          * shake your hand personally. :)
1139          */
1140         foreach ($bad_attvals as $matchtag=>$matchattrs){
1141             if (preg_match($matchtag, $tagname)){
1142                 foreach ($matchattrs as $matchattr=>$valary){
1143                     if (preg_match($matchattr, $attname)){
1144                         /**
1145                          * There are two arrays in valary.
1146                          * First is matches.
1147                          * Second one is replacements
1148                          */
1149                         list($valmatch, $valrepl) = $valary;
1150                         $newvalue =
1151                             preg_replace($valmatch, $valrepl, $attvalue);
1152                         if ($newvalue != $attvalue){
1153                             $attary{$attname} = $newvalue;
1154                         }
1155                     }
1156                 }
1157             }
1158         }
1159         /**
1160          * Turn cid: urls into http-friendly ones.
1161          */
1162         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1163             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1164         }
1165     }
1166     /**
1167      * See if we need to append any attributes to this tag.
1168      */
1169     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1170         if (preg_match($matchtag, $tagname)){
1171             $attary = array_merge($attary, $addattary);
1172         }
1173     }
1174     return $attary;
1175 }
1176
1177 /**
1178  * This function edits the style definition to make them friendly and
1179  * usable in squirrelmail.
1180  *
1181  * @param  $message  the message object
1182  * @param  $id       the message id
1183  * @param  $content  a string with whatever is between <style> and </style>
1184  * @return           a string with edited content.
1185  */
1186 function sq_fixstyle($message, $id, $content){
1187     global $view_unsafe_images;
1188     $me = 'sq_fixstyle';
1189     /**
1190      * First look for general BODY style declaration, which would be
1191      * like so:
1192      * body {background: blah-blah}
1193      * and change it to .bodyclass so we can just assign it to a <div>
1194      */
1195     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1196     $secremoveimg = "../images/" . _("sec_remove_eng.png");
1197     /**
1198      * Fix url('blah') declarations.
1199      */
1200     $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1201                             "url(\\1$secremoveimg\\2)", $content);
1202     /**
1203      * Fix url('https*://.*) declarations but only if $view_unsafe_images
1204      * is false.
1205      */
1206     if (!$view_unsafe_images){
1207         $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1208                                 "url(\\1$secremoveimg\\2)", $content);
1209     }
1210
1211     /**
1212      * Fix urls that refer to cid:
1213      */
1214     while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1215                       $matches)){
1216         $cidurl = $matches{1};
1217         $httpurl = sq_cid2http($message, $id, $cidurl);
1218         $content = preg_replace("|url\($cidurl\)|si",
1219                                 "url($httpurl)", $content);
1220     }
1221
1222     /**
1223      * Fix stupid css declarations which lead to vulnerabilities
1224      * in IE.
1225      */
1226     $match   = Array('/expression/si',
1227                      '/behaviou*r/si',
1228                      '/binding/si');
1229     $replace = Array('idiocy', 'idiocy', 'idiocy');
1230     $content = preg_replace($match, $replace, $content);
1231     return $content;
1232 }
1233
1234 /**
1235  * This function converts cid: url's into the ones that can be viewed in
1236  * the browser.
1237  *
1238  * @param  $message  the message object
1239  * @param  $id       the message id
1240  * @param  $cidurl   the cid: url.
1241  * @return           a string with a http-friendly url
1242  */
1243 function sq_cid2http($message, $id, $cidurl, $mailbox){
1244     /**
1245      * Get rid of quotes.
1246      */
1247     $quotchar = substr($cidurl, 0, 1);
1248     $cidurl = str_replace($quotchar, "", $cidurl);
1249     $cidurl = substr(trim($cidurl), 4);
1250     $linkurl = find_ent_id($cidurl, $message);
1251     /* in case of non-save cid links $httpurl should be replaced by a sort of
1252        unsave link image */
1253     $httpurl = '';
1254     if ($linkurl) {
1255         $httpurl = $quotchar . "../src/download.php?absolute_dl=true&amp;" .
1256                    "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1257                    "&amp;ent_id=" . $linkurl . $quotchar;
1258     }
1259     return $httpurl;
1260 }
1261
1262 /**
1263  * This function changes the <body> tag into a <div> tag since we
1264  * can't really have a body-within-body.
1265  *
1266  * @param  $attary  an array of attributes and values of <body>
1267  * @return          a modified array of attributes to be set for <div>
1268  */
1269 function sq_body2div($attary){
1270     $me = 'sq_body2div';
1271     $divattary = Array( 'class' => "'bodyclass'" );
1272     $bgcolor = '#ffffff';
1273     $text = '#000000';
1274     $styledef = '';
1275     if (is_array($attary) && sizeof($attary) > 0){
1276         foreach ($attary as $attname=>$attvalue){
1277             $quotchar = substr($attvalue, 0, 1);
1278             $attvalue = str_replace($quotchar, "", $attvalue);
1279             switch ($attname){
1280             case 'background':
1281                 $styledef .= "background-image: url('$attvalue'); ";
1282                 break;
1283             case 'bgcolor':
1284                 $styledef .= "background-color: $attvalue; ";
1285                 break;
1286             case 'text':
1287                 $styledef .= "color: $attvalue; ";
1288             }
1289         }
1290         if (strlen($styledef) > 0){
1291             $divattary{"style"} = "\"$styledef\"";
1292         }
1293     }
1294     return $divattary;
1295 }
1296
1297 /**
1298  * This is the main function and the one you should actually be calling.
1299  * There are several variables you should be aware of an which need
1300  * special description.
1301  *
1302  * Since the description is quite lengthy, see it here:
1303  * http://www.mricon.com/html/phpfilter.html
1304  *
1305  * @param $body                 the string with HTML you wish to filter
1306  * @param $tag_list             see description above
1307  * @param $rm_tags_with_content see description above
1308  * @param $self_closing_tags    see description above
1309  * @param $force_tag_closing    see description above
1310  * @param $rm_attnames          see description above
1311  * @param $bad_attvals          see description above
1312  * @param $add_attr_to_tag      see description above
1313  * @param $message              message object
1314  * @param $id                   message id
1315  * @return                      sanitized html safe to show on your pages.
1316  */
1317 function sq_sanitize($body,
1318                      $tag_list,
1319                      $rm_tags_with_content,
1320                      $self_closing_tags,
1321                      $force_tag_closing,
1322                      $rm_attnames,
1323                      $bad_attvals,
1324                      $add_attr_to_tag,
1325                      $message,
1326                      $id,
1327                      $mailbox
1328                      ){
1329     $me = 'sq_sanitize';
1330     /**
1331      * Normalize rm_tags and rm_tags_with_content.
1332      */
1333     @array_walk($rm_tags, 'sq_casenormalize');
1334     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1335     @array_walk($self_closing_tags, 'sq_casenormalize');
1336     /**
1337      * See if tag_list is of tags to remove or tags to allow.
1338      * false  means remove these tags
1339      * true   means allow these tags
1340      */
1341     $rm_tags = array_shift($tag_list);
1342     $curpos = 0;
1343     $open_tags = Array();
1344     $trusted = "<!-- begin sanitized html -->\n";
1345     $skip_content = false;
1346     /**
1347      * Take care of netscape's stupid javascript entities like
1348      * &{alert('boo')};
1349      */
1350     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1351
1352     while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1353         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1354         $free_content = substr($body, $curpos, $lt-$curpos);
1355         /**
1356          * Take care of <style>
1357          */
1358         if ($tagname == "style" && $tagtype == 2){
1359             /**
1360              * This is a closing </style>. Edit the
1361              * content before we apply it.
1362              */
1363             $free_content = sq_fixstyle($message, $id, $free_content);
1364         }
1365         if ($skip_content == false){
1366             $trusted .= $free_content;
1367         } else {
1368         }
1369         if ($tagname != FALSE){
1370             if ($tagtype == 2){
1371                 if ($skip_content == $tagname){
1372                     /**
1373                      * Got to the end of tag we needed to remove.
1374                      */
1375                     $tagname = false;
1376                     $skip_content = false;
1377                 } else {
1378                     if ($skip_content == false){
1379                         if ($tagname == "body"){
1380                             $tagname = "div";
1381                         } else {
1382                             if (isset($open_tags{$tagname}) &&
1383                                 $open_tags{$tagname} > 0){
1384                                 $open_tags{$tagname}--;
1385                             } else {
1386                                 $tagname = false;
1387                             }
1388                         }
1389                     } else {
1390                     }
1391                 }
1392             } else {
1393                 /**
1394                  * $rm_tags_with_content
1395                  */
1396                 if ($skip_content == false){
1397                     /**
1398                      * See if this is a self-closing type and change
1399                      * tagtype appropriately.
1400                      */
1401                     if ($tagtype == 1
1402                         && in_array($tagname, $self_closing_tags)){
1403                         $tagtype=3;
1404                     }
1405                     /**
1406                      * See if we should skip this tag and any content
1407                      * inside it.
1408                      */
1409                     if ($tagtype == 1 &&
1410                         in_array($tagname, $rm_tags_with_content)){
1411                         $skip_content = $tagname;
1412                     } else {
1413                         if (($rm_tags == false
1414                              && in_array($tagname, $tag_list)) ||
1415                             ($rm_tags == true &&
1416                              !in_array($tagname, $tag_list))){
1417                             $tagname = false;
1418                         } else {
1419                             if ($tagtype == 1){
1420                                 if (isset($open_tags{$tagname})){
1421                                     $open_tags{$tagname}++;
1422                                 } else {
1423                                     $open_tags{$tagname}=1;
1424                                 }
1425                             }
1426                             /**
1427                              * This is where we run other checks.
1428                              */
1429                             if (is_array($attary) && sizeof($attary) > 0){
1430                                 $attary = sq_fixatts($tagname,
1431                                                      $attary,
1432                                                      $rm_attnames,
1433                                                      $bad_attvals,
1434                                                      $add_attr_to_tag,
1435                                                      $message,
1436                                                      $id,
1437                                                      $mailbox
1438                                                      );
1439                             }
1440                             /**
1441                              * Convert body into div.
1442                              */
1443                             if ($tagname == "body"){
1444                                 $tagname = "div";
1445                                 $attary = sq_body2div($attary, $message, $id);
1446                             }
1447                         }
1448                     }
1449                 } else {
1450                 }
1451             }
1452             if ($tagname != false && $skip_content == false){
1453                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1454             }
1455         } else {
1456         }
1457         $curpos = $gt+1;
1458     }
1459     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1460     if ($force_tag_closing == true){
1461         foreach ($open_tags as $tagname=>$opentimes){
1462             while ($opentimes > 0){
1463                 $trusted .= '</' . $tagname . '>';
1464                 $opentimes--;
1465             }
1466         }
1467         $trusted .= "\n";
1468     }
1469     $trusted .= "<!-- end sanitized html -->\n";
1470     return $trusted;
1471 }
1472
1473 /**
1474  * This is a wrapper function to call html sanitizing routines.
1475  *
1476  * @param  $body  the body of the message
1477  * @param  $id    the id of the message
1478  * @return        a string with html safe to display in the browser.
1479  */
1480 function magicHTML($body, $id, $message, $mailbox = 'INBOX'){
1481     global $attachment_common_show_images, $view_unsafe_images,
1482         $has_unsafe_images;
1483     /**
1484      * Don't display attached images in HTML mode.
1485      */
1486     $attachment_common_show_images = false;
1487     $tag_list = Array(
1488                       false,
1489                       "object",
1490                       "meta",
1491                       "html",
1492                       "head",
1493                       "base",
1494                       "link"
1495                       );
1496
1497     $rm_tags_with_content = Array(
1498                                   "script",
1499                                   "applet",
1500                                   "embed",
1501                                   "title"
1502                                   );
1503
1504     $self_closing_tags =  Array(
1505                                 "img",
1506                                 "br",
1507                                 "hr",
1508                                 "input"
1509                                 );
1510
1511     $force_tag_closing = false;
1512
1513     $rm_attnames = Array(
1514                          "/.*/" =>
1515                          Array(
1516                                "/target/si",
1517                                "/^on.*/si",
1518                                "/^dynsrc/si",
1519                                "/^data.*/si"
1520                                )
1521                          );
1522
1523     $secremoveimg = "../images/" . _("sec_remove_eng.png");
1524     $bad_attvals = Array(
1525         "/.*/" =>
1526             Array(
1527                 "/^src|background/i" =>
1528                     Array(
1529                           Array(
1530                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1531                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1532                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1533                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1534                                 ),
1535                           Array(
1536                                 "\\1$secremoveimg\\2",
1537                                 "\\1$secremoveimg\\2",
1538                                 "\\1$secremoveimg\\2",
1539                                 "\\1$secremoveimg\\2"
1540                                 )
1541                         ),
1542                 "/^href|action/i" =>
1543                     Array(
1544                           Array(
1545                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1546                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1547                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1548                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1549                                 ),
1550                           Array(
1551                                 "\\1#\\2",
1552                                 "\\1#\\2",
1553                                 "\\1#\\2",
1554                                 "\\1#\\2"
1555                                 )
1556                         ),
1557                 "/^style/si" =>
1558                     Array(
1559                           Array(
1560                                 "/expression/si",
1561                                 "/binding/si",
1562                                 "/behaviou*r/si",
1563                                 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1564                                 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1565                                 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1566                                 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1567                                ),
1568                           Array(
1569                                 "idiocy",
1570                                 "idiocy",
1571                                 "idiocy",
1572                                 "url(\\1#\\2)",
1573                                 "url(\\1#\\2)",
1574                                 "url(\\1#\\2)",
1575                                 "url(\\1#\\2)"
1576                                )
1577                           )
1578                 )
1579         );
1580     if (!$view_unsafe_images){
1581         /**
1582          * Remove any references to http/https if view_unsafe_images set
1583          * to false.
1584          */
1585          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1586                     '/^([\'\"])\s*https*:.*([\'\"])/si');
1587          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1588                     "\\1$secremoveimg\\2");
1589          array_push($bad_attvals{'/.*/'}{'/^style/si'}[0],
1590                     '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1591          array_push($bad_attvals{'/.*/'}{'/^style/si'}[1],
1592                     "url(\\1$secremoveimg\\2)");
1593     }
1594
1595     $add_attr_to_tag = Array(
1596                              "/^a$/si" => Array('target'=>'"_new"')
1597                              );
1598     $trusted = sq_sanitize($body,
1599                            $tag_list,
1600                            $rm_tags_with_content,
1601                            $self_closing_tags,
1602                            $force_tag_closing,
1603                            $rm_attnames,
1604                            $bad_attvals,
1605                            $add_attr_to_tag,
1606                            $message,
1607                            $id,
1608                            $mailbox
1609                            );
1610     if (preg_match("|$secremoveimg|si", $trusted)){
1611         $has_unsafe_images = true;
1612     }
1613     return $trusted;
1614 }
1615
1616 ?>