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