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