functions/mime.php

   1 <?php
   2
   3 /**
   4  * mime.php
   5  *
   6  * Copyright (c) 1999-2003 The SquirrelMail Project Team
   7  * Licensed under the GNU GPL. For full terms see the file COPYING.
   8  *
   9  * This contains the functions necessary to detect and decode MIME
  10  * messages.
  11  *
  12  * $Id$
  13  */
  14
  15 require_once(SM_PATH . 'functions/imap.php');
  16 require_once(SM_PATH . '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     $read = trim(substr ($read, 0, -1));
  32     $i = 0;
  33     $msg = Message::parseStructure($read,$i);
  34     if (!is_object($msg)) {
  35         include_once(SM_PATH . 'functions/display_messages.php');
  36         global $color, $mailbox;
  37         displayPageHeader( $color, urldecode($mailbox) );
  38         echo "<BODY TEXT=\"$color[8]\" BGCOLOR=\"$color[4]\" LINK=\"$color[7]\" VLINK=\"$color[7]\" ALINK=\"$color[7]\">\n\n" .
  39          '<CENTER>';
  40         $errormessage  = _("SquirrelMail could not decode the bodystructure of the message");
  41         $errormessage .= '<BR>'._("the provided bodystructure by your imap-server").':<BR><BR>';
  42         $errormessage .= '<table><tr><td>' . htmlspecialchars($read) . '</td></tr></table>';
  43         plain_error_message( $errormessage, $color );
  44         echo '</body></html>';
  45         exit;
  46     }
  47     if (count($flags)) {
  48         foreach ($flags as $flag) {
  49             $char = strtoupper($flag{1});
  50             switch ($char) {
  51                 case 'S':
  52                     if (strtolower($flag) == '\\seen') {
  53                         $msg->is_seen = true;
  54                     }
  55                     break;
  56                 case 'A':
  57                     if (strtolower($flag) == '\\answered') {
  58                         $msg->is_answered = true;
  59                     }
  60                     break;
  61                 case 'D':
  62                     if (strtolower($flag) == '\\deleted') {
  63                         $msg->is_deleted = true;
  64                     }
  65                     break;
  66                 case 'F':
  67                     if (strtolower($flag) == '\\flagged') {
  68                         $msg->is_flagged = true;
  69                     }
  70                     break;
  71                 case 'M':
  72                     if (strtolower($flag) == '$mdnsent') {
  73                         $msg->is_mdnsent = true;
  74                     }
  75                     break;
  76                 default:
  77                     break;
  78             }
  79         }
  80     }
  81     //    listEntities($msg);
  82     return $msg;
  83 }
  84
  85
  86
  87 /* This starts the parsing of a particular structure.  It is called recursively,
  88  * so it can be passed different structures.  It returns an object of type
  89  * $message.
  90  * First, it checks to see if it is a multipart message.  If it is, then it
  91  * handles that as it sees is necessary.  If it is just a regular entity,
  92  * then it parses it and adds the necessary header information (by calling out
  93  * to mime_get_elements()
  94  */
  95
  96 function mime_fetch_body($imap_stream, $id, $ent_id=1) {
  97     global $uid_support;
  98     /* Do a bit of error correction.  If we couldn't find the entity id, just guess
  99      * that it is the first one.  That is usually the case anyway.
 100      */
 101     if (!$ent_id) {
 102         $cmd = "FETCH $id BODY[]";
 103     } else {
 104         $cmd = "FETCH $id BODY[$ent_id]";
 105     }
 106
 107     $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, $uid_support);
 108     do {
 109         $topline = trim(array_shift($data));
 110     } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
 111
 112     $wholemessage = implode('', $data);
 113     if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
 114         $ret = substr($wholemessage, 0, $regs[1]);
 115         /* There is some information in the content info header that could be important
 116          * in order to parse html messages. Let's get them here.
 117          */
 118         if ($ret{0} == '<') {
 119             $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, $uid_support);
 120         }
 121     } else if (ereg('"([^"]*)"', $topline, $regs)) {
 122         $ret = $regs[1];
 123     } else {
 124         global $where, $what, $mailbox, $passed_id, $startMessage;
 125         $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
 126         if (isset($where) && isset($what)) {
 127             $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
 128         } else {
 129             $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 130         }
 131         $par .= '&amp;response=' . urlencode($response) .
 132                 '&amp;message='  . urlencode($message)  .
 133                 '&amp;topline='  . urlencode($topline);
 134
 135         echo   '<tt><br>' .
 136                '<table width="80%"><tr>' .
 137                '<tr><td colspan=2>' .
 138                _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
 139                '</td></tr>' .
 140                '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
 141                '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
 142                '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
 143                '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
 144                "</table><BR></tt></font><hr>";
 145
 146         $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, $uid_support);
 147         array_shift($data);
 148         $wholemessage = implode('', $data);
 149
 150         $ret = $wholemessage;
 151     }
 152     return $ret;
 153 }
 154
 155 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding) {
 156     global $uid_support;
 157
 158     $sid = sqimap_session_id($uid_support);
 159     /* Don't kill the connection if the browser is over a dialup
 160      * and it would take over 30 seconds to download it.
 161      * Don´t call set_time_limit in safe mode.
 162      */
 163
 164     if (!ini_get('safe_mode')) {
 165         set_time_limit(0);
 166     }
 167     if ($uid_support) {
 168        $sid_s = substr($sid,0,strpos($sid, ' '));
 169     } else {
 170        $sid_s = $sid;
 171     }
 172
 173     $body = mime_fetch_body ($imap_stream, $id, $ent_id);
 174     echo decodeBody($body, $encoding);
 175     return;
 176 /*
 177     fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
 178     $cnt = 0;
 179     $continue = true;
 180     $read = fgets ($imap_stream,8192);
 181
 182
 183     // This could be bad -- if the section has sqimap_session_id() . ' OK'
 184     // or similar, it will kill the download.
 185     while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
 186         if (trim($read) == ')==') {
 187             $read1 = $read;
 188             $read = fgets ($imap_stream,4096);
 189             if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
 190                 return;
 191             } else {
 192                 echo decodeBody($read1, $encoding) .
 193                      decodeBody($read, $encoding);
 194             }
 195         } else if ($cnt) {
 196             echo decodeBody($read, $encoding);
 197         }
 198         $read = fgets ($imap_stream,4096);
 199         $cnt++;
 200 //      break;
 201     }
 202 */
 203 }
 204
 205 /* -[ END MIME DECODING ]----------------------------------------------------------- */
 206
 207 /* This is here for debugging purposes.  It will print out a list
 208  * of all the entity IDs that are in the $message object.
 209  */
 210 function listEntities ($message) {
 211     if ($message) {
 212         echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br>';
 213         for ($i = 0; isset($message->entities[$i]); $i++) {
 214             echo "$i : ";
 215             $msg = listEntities($message->entities[$i]);
 216
 217             if ($msg) {
 218                 echo "return: ";
 219                 return $msg;
 220             }
 221         }
 222     }
 223 }
 224
 225 function getPriorityStr($priority) {
 226     $priority_level = substr($priority,0,1);
 227
 228     switch($priority_level) {
 229         /* Check for a higher then normal priority. */
 230         case '1':
 231         case '2':
 232             $priority_string = _("High");
 233             break;
 234
 235         /* Check for a lower then normal priority. */
 236         case '4':
 237         case '5':
 238             $priority_string = _("Low");
 239             break;
 240
 241         /* Check for a normal priority. */
 242         case '3':
 243         default:
 244             $priority_level = '3';
 245             $priority_string = _("Normal");
 246             break;
 247
 248     }
 249     return $priority_string;
 250 }
 251
 252 /* returns a $message object for a particular entity id */
 253 function getEntity ($message, $ent_id) {
 254     return $message->getEntity($ent_id);
 255 }
 256
 257 /* translateText
 258  * Extracted from strings.php 23/03/2002
 259  */
 260
 261 function translateText(&$body, $wrap_at, $charset) {
 262     global $where, $what;   /* from searching */
 263     global $color;          /* color theme */
 264
 265     require_once(SM_PATH . 'functions/url_parser.php');
 266
 267     $body_ary = explode("\n", $body);
 268     for ($i=0; $i < count($body_ary); $i++) {
 269         $line = $body_ary[$i];
 270         if (strlen($line) - 2 >= $wrap_at) {
 271             sqWordWrap($line, $wrap_at);
 272         }
 273         $line = charset_decode($charset, $line);
 274         $line = str_replace("\t", '        ', $line);
 275
 276         parseUrl ($line);
 277
 278         $quotes = 0;
 279         $pos = 0;
 280         $j = strlen($line);
 281
 282         while ($pos < $j) {
 283             if ($line[$pos] == ' ') {
 284                 $pos++;
 285             } else if (strpos($line, '&gt;', $pos) === $pos) {
 286                 $pos += 4;
 287                 $quotes++;
 288             } else {
 289                 break;
 290             }
 291         }
 292
 293         if ($quotes > 1) {
 294             if (!isset($color[14])) {
 295                 $color[14] = '#FF0000';
 296             }
 297             $line = '<FONT COLOR="' . $color[14] . '">' . $line . '</FONT>';
 298         } elseif ($quotes) {
 299             if (!isset($color[13])) {
 300                 $color[13] = '#800000';
 301             }
 302             $line = '<FONT COLOR="' . $color[13] . '">' . $line . '</FONT>';
 303         }
 304
 305         $body_ary[$i] = $line;
 306     }
 307     $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
 308 }
 309
 310
 311 /* This returns a parsed string called $body. That string can then
 312  * be displayed as the actual message in the HTML. It contains
 313  * everything needed, including HTML Tags, Attachments at the
 314  * bottom, etc.
 315  */
 316 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
 317     /* This if statement checks for the entity to show as the
 318      * primary message. To add more of them, just put them in the
 319      * order that is their priority.
 320      */
 321     global $startMessage, $username, $key, $imapServerAddress, $imapPort,
 322            $show_html_default, $sort, $has_unsafe_images, $passed_ent_id;
 323
 324     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
 325         $view_unsafe_images = false;
 326     }
 327
 328     $body = '';
 329     $urlmailbox = urlencode($mailbox);
 330     $body_message = getEntity($message, $ent_num);
 331     if (($body_message->header->type0 == 'text') ||
 332         ($body_message->header->type0 == 'rfc822')) {
 333         $body = mime_fetch_body ($imap_stream, $id, $ent_num);
 334         $body = decodeBody($body, $body_message->header->encoding);
 335         $hookResults = do_hook("message_body", $body);
 336         $body = $hookResults[1];
 337
 338         /* If there are other types that shouldn't be formatted, add
 339          * them here.
 340          */
 341
 342         if ($body_message->header->type1 == 'html') {
 343             if ($show_html_default <> 1) {
 344                 $entity_conv = array('&nbsp;' => ' ',
 345                                      '<p>'    => "\n",
 346                                      '<br>'   => "\n",
 347                                      '<P>'    => "\n",
 348                                      '<BR>'   => "\n",
 349                                      '&gt;'   => '>',
 350                                      '&lt;'   => '<');
 351                 $body = strtr($body, $entity_conv);
 352                 $body = strip_tags($body);
 353                 $body = trim($body);
 354                 translateText($body, $wrap_at,
 355                               $body_message->header->getParameter('charset'));
 356             } else {
 357                 $body = magicHTML($body, $id, $message, $mailbox);
 358             }
 359         } else {
 360             translateText($body, $wrap_at,
 361                           $body_message->header->getParameter('charset'));
 362         }
 363         $link = 'read_body.php?passed_id=' . $id . '&amp;ent_id='.$ent_num.
 364                  '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
 365                  '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 366         if (isset($passed_ent_id)) {
 367             $link .= '&amp;passed_ent_id='.$passed_ent_id;
 368         }
 369         if ($view_unsafe_images) {
 370             $text = _("Hide Unsafe Images");
 371         } else {
 372             if (isset($has_unsafe_images) && $has_unsafe_images) {
 373                 $link .= '&amp;view_unsafe_images=1';
 374                 $text = _("View Unsafe Images");
 375             } else {
 376                 $text = '';
 377             }
 378         }
 379         $body .= '<center><small><a href="'.$link.'">'.$text.
 380                  '</a></small></center><br>' . "\n";
 381     }
 382     return $body;
 383 }
 384
 385
 386 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 387     global $where, $what, $startMessage, $color;
 388     static $ShownHTML = 0;
 389
 390     $att_ar = $message->getAttachments($exclude_id);
 391
 392     if (!count($att_ar)) return '';
 393
 394     $attachments = '';
 395
 396     $urlMailbox = urlencode($mailbox);
 397
 398     foreach ($att_ar as $att) {
 399         $ent = urldecode($att->entity_id);
 400         $header = $att->header;
 401         $type0 = strtolower($header->type0);
 402         $type1 = strtolower($header->type1);
 403         $name = '';
 404         $links['download link']['text'] = _("download");
 405         $links['download link']['href'] =
 406                 "../src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 407         $ImageURL = '';
 408         if ($type0 =='message' && $type1 == 'rfc822') {
 409             $default_page = '../src/read_body.php';
 410             $rfc822_header = $att->rfc822_header;
 411             $filename = decodeHeader($rfc822_header->subject);
 412             if (trim( $filename ) == '') {
 413                 $filename = 'untitled-[' . $ent . ']' ;
 414             }
 415             $from_o = $rfc822_header->from;
 416             if (is_object($from_o)) {
 417                 $from_name = $from_o->getAddress(false);
 418             } else {
 419                 $from_name = _("Unknown sender");
 420             }
 421             $from_name = decodeHeader(htmlspecialchars($from_name));
 422             $description = $from_name;
 423         } else {
 424             $default_page = '../src/download.php';
 425             if (is_object($header->disposition)) {
 426                 $filename = decodeHeader($header->disposition->getProperty('filename'));
 427                 if (trim($filename) == '') {
 428                     $name = decodeHeader($header->disposition->getProperty('name'));
 429                     if (trim($name) == '') {
 430                         $name = decodeHeader($header->getParameter('name'));
 431                         if(trim($name) == '') {
 432                             if (trim( $header->id ) == '') {
 433                                 $filename = 'untitled-[' . $ent . ']' ;
 434                             } else {
 435                                 $filename = 'cid: ' . $header->id;
 436                             }
 437                         } else {
 438                             $filename = $name;
 439                         }
 440                     } else {
 441                         $filename = $name;
 442                     }
 443                 }
 444             } else {
 445                 if (trim( $header->id ) == '') {
 446                     $filename = 'untitled-[' . $ent . ']' ;
 447                 } else {
 448                     $filename = 'cid: ' . $header->id;
 449                 }
 450             }
 451
 452             if ($header->description) {
 453                 $description = htmlspecialchars($header->description);
 454             } else {
 455                 $description = '';
 456             }
 457         }
 458
 459         $display_filename = $filename;
 460         if (isset($passed_ent_id)) {
 461             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 462         } else {
 463             $passed_ent_id_link = '';
 464         }
 465         $defaultlink = $default_page . "?startMessage=$startMessage"
 466                      . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 467                      . '&amp;ent_id='.$ent.$passed_ent_id_link.'&amp;absolute_dl=true';
 468         if ($where && $what) {
 469            $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 470         }
 471         /* This executes the attachment hook with a specific MIME-type.
 472          * If that doesn't have results, it tries if there's a rule
 473          * for a more generic type.
 474          */
 475         $hookresults = do_hook("attachment $type0/$type1", $links,
 476                                $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 477                                $display_filename, $where, $what);
 478         if(count($hookresults[1]) <= 1) {
 479             $hookresults = do_hook("attachment $type0/*", $links,
 480                                    $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 481                                    $display_filename, $where, $what);
 482         }
 483
 484         $links = $hookresults[1];
 485         $defaultlink = $hookresults[6];
 486
 487         $attachments .= '<TR><TD>' .
 488                         "<A HREF=\"$defaultlink\">$display_filename</A>&nbsp;</TD>" .
 489                         '<TD><SMALL><b>' . show_readable_size($header->size) .
 490                         '</b>&nbsp;&nbsp;</small></TD>' .
 491                         "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
 492                         '<TD><SMALL>';
 493         $attachments .= '<b>' . $description . '</b>';
 494         $attachments .= '</SMALL></TD><TD><SMALL>&nbsp;';
 495
 496         $skipspaces = 1;
 497         foreach ($links as $val) {
 498             if ($skipspaces) {
 499                 $skipspaces = 0;
 500             } else {
 501                 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
 502             }
 503             $attachments .= '<a href="' . $val['href'] . '">' .  $val['text'] . '</a>';
 504         }
 505         unset($links);
 506         $attachments .= "</TD></TR>\n";
 507     }
 508     return $attachments;
 509 }
 510
 511 /* This function decodes the body depending on the encoding type. */
 512 function decodeBody($body, $encoding) {
 513     global $languages, $squirrelmail_language;
 514     global $show_html_default;
 515
 516     $body = str_replace("\r\n", "\n", $body);
 517     $encoding = strtolower($encoding);
 518
 519     if ($encoding == 'quoted-printable' ||
 520         $encoding == 'quoted_printable') {
 521         $body = quoted_printable_decode($body);
 522
 523         while (ereg("=\n", $body)) {
 524             $body = ereg_replace ("=\n", '', $body);
 525         }
 526
 527     } else if ($encoding == 'base64') {
 528         $body = base64_decode($body);
 529     }
 530
 531     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 532         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 533         $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
 534     }
 535
 536     // All other encodings are returned raw.
 537     return $body;
 538 }
 539
 540 /*
 541  * This functions decode strings that is encoded according to
 542  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 543  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 544  */
 545 function decodeHeader ($string, $utfencode=true) {
 546     global $languages, $squirrelmail_language;
 547     if (is_array($string)) {
 548         $string = implode("\n", $string);
 549     }
 550
 551     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 552         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 553         $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
 554     }
 555
 556     $i = 0;
 557     while (preg_match('/^(.{' . $i . '})(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=/Ui',
 558                       $string, $res)) {
 559         $prefix = $res[1];
 560         /* Ignore white-space between consecutive encoded-words. */
 561         if (strspn($res[2], " \t") != strlen($res[2])) {
 562             $prefix .= $res[2];
 563         }
 564
 565         if (ucfirst($res[4]) == 'B') {
 566             $replace = base64_decode($res[5]);
 567         } else {
 568             $replace = str_replace('_', ' ', $res[5]);
 569             $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 570                                     $replace);
 571             /* Only encode into entities by default. Some places
 572              * don't need the encoding, like the compose form.
 573              */
 574             if ($utfencode) {
 575                 $replace = charset_decode($res[3], $replace);
 576             }
 577         }
 578         $string = $prefix . $replace . substr($string, strlen($res[0]));
 579         $i = strlen($prefix) + strlen($replace);
 580     }
 581     return $string;
 582 }
 583
 584 /*
 585  * Encode a string according to RFC 1522 for use in headers if it
 586  * contains 8-bit characters or anything that looks like it should
 587  * be encoded.
 588  */
 589 function encodeHeader ($string) {
 590     global $default_charset, $languages, $squirrelmail_language;
 591
 592     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 593         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 594         return  $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
 595     }
 596
 597     // Encode only if the string contains 8-bit characters or =?
 598     $j = strlen($string);
 599     $l = strstr($string, '=?');         // Must be encoded ?
 600     $ret = '';
 601     for($i = 0; $i < $j; ++$i) {
 602         switch($string{$i}) {
 603             case '=':
 604                 $ret .= '=3D';
 605                 break;
 606             case '?':
 607                 $ret .= '=3F';
 608                 break;
 609             case '_':
 610                 $ret .= '=5F';
 611                 break;
 612             case ' ':
 613                 $ret .= '_';
 614                 break;
 615             default:
 616                 $k = ord($string{$i});
 617                 if ($k > 126) {
 618                     $ret .= sprintf("=%02X", $k);
 619                     $l = TRUE;
 620                 } else {
 621                     $ret .= $string{$i};
 622                 }
 623                 break;
 624         }
 625     }
 626
 627     if ($l) {
 628         $string = "=?$default_charset?Q?$ret?=";
 629     }
 630
 631     return $string;
 632 }
 633
 634 /* This function trys to locate the entity_id of a specific mime element */
 635 function find_ent_id($id, $message) {
 636     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
 637         if ($message->entities[$i]->header->type0 == 'multipart')  {
 638             $ret = find_ent_id($id, $message->entities[$i]);
 639         } else {
 640             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
 641 //                if (sq_check_save_extension($message->entities[$i])) {
 642                     return $message->entities[$i]->entity_id;
 643 //                }
 644             }
 645         }
 646     }
 647     return $ret;
 648 }
 649
 650 function sq_check_save_extension($message) {
 651     $filename = $message->getFilename();
 652     $ext = substr($filename, strrpos($filename,'.')+1);
 653     $save_extensions = array('jpg','jpeg','gif','png','bmp');
 654     return in_array($ext, $save_extensions);
 655 }
 656
 657
 658 /**
 659  ** HTMLFILTER ROUTINES
 660  */
 661
 662 /**
 663  * This function returns the final tag out of the tag name, an array
 664  * of attributes, and the type of the tag. This function is called by
 665  * sq_sanitize internally.
 666  *
 667  * @param  $tagname  the name of the tag.
 668  * @param  $attary   the array of attributes and their values
 669  * @param  $tagtype  The type of the tag (see in comments).
 670  * @return           a string with the final tag representation.
 671  */
 672 function sq_tagprint($tagname, $attary, $tagtype){
 673     $me = 'sq_tagprint';
 674
 675     if ($tagtype == 2){
 676         $fulltag = '</' . $tagname . '>';
 677     } else {
 678         $fulltag = '<' . $tagname;
 679         if (is_array($attary) && sizeof($attary)){
 680             $atts = Array();
 681             while (list($attname, $attvalue) = each($attary)){
 682                 array_push($atts, "$attname=$attvalue");
 683             }
 684             $fulltag .= ' ' . join(" ", $atts);
 685         }
 686         if ($tagtype == 3){
 687             $fulltag .= ' /';
 688         }
 689         $fulltag .= '>';
 690     }
 691     return $fulltag;
 692 }
 693
 694 /**
 695  * A small helper function to use with array_walk. Modifies a by-ref
 696  * value and makes it lowercase.
 697  *
 698  * @param  $val a value passed by-ref.
 699  * @return      void since it modifies a by-ref value.
 700  */
 701 function sq_casenormalize(&$val){
 702     $val = strtolower($val);
 703 }
 704
 705 /**
 706  * This function skips any whitespace from the current position within
 707  * a string and to the next non-whitespace value.
 708  *
 709  * @param  $body   the string
 710  * @param  $offset the offset within the string where we should start
 711  *                 looking for the next non-whitespace character.
 712  * @return         the location within the $body where the next
 713  *                 non-whitespace char is located.
 714  */
 715 function sq_skipspace($body, $offset){
 716     $me = 'sq_skipspace';
 717     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
 718     if (sizeof($matches{1})){
 719         $count = strlen($matches{1});
 720         $offset += $count;
 721     }
 722     return $offset;
 723 }
 724
 725 /**
 726  * This function looks for the next character within a string.  It's
 727  * really just a glorified "strpos", except it catches if failures
 728  * nicely.
 729  *
 730  * @param  $body   The string to look for needle in.
 731  * @param  $offset Start looking from this position.
 732  * @param  $needle The character/string to look for.
 733  * @return         location of the next occurance of the needle, or
 734  *                 strlen($body) if needle wasn't found.
 735  */
 736 function sq_findnxstr($body, $offset, $needle){
 737     $me  = 'sq_findnxstr';
 738     $pos = strpos($body, $needle, $offset);
 739     if ($pos === FALSE){
 740         $pos = strlen($body);
 741     }
 742     return $pos;
 743 }
 744
 745 /**
 746  * This function takes a PCRE-style regexp and tries to match it
 747  * within the string.
 748  *
 749  * @param  $body   The string to look for needle in.
 750  * @param  $offset Start looking from here.
 751  * @param  $reg    A PCRE-style regex to match.
 752  * @return         Returns a false if no matches found, or an array
 753  *                 with the following members:
 754  *                 - integer with the location of the match within $body
 755  *                 - string with whatever content between offset and the match
 756  *                 - string with whatever it is we matched
 757  */
 758 function sq_findnxreg($body, $offset, $reg){
 759     $me = 'sq_findnxreg';
 760     $matches = Array();
 761     $retarr = Array();
 762     preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
 763     if (!$matches{0}){
 764         $retarr = false;
 765     } else {
 766         $retarr{0} = $offset + strlen($matches{1});
 767         $retarr{1} = $matches{1};
 768         $retarr{2} = $matches{2};
 769     }
 770     return $retarr;
 771 }
 772
 773 /**
 774  * This function looks for the next tag.
 775  *
 776  * @param  $body   String where to look for the next tag.
 777  * @param  $offset Start looking from here.
 778  * @return         false if no more tags exist in the body, or
 779  *                 an array with the following members:
 780  *                 - string with the name of the tag
 781  *                 - array with attributes and their values
 782  *                 - integer with tag type (1, 2, or 3)
 783  *                 - integer where the tag starts (starting "<")
 784  *                 - integer where the tag ends (ending ">")
 785  *                 first three members will be false, if the tag is invalid.
 786  */
 787 function sq_getnxtag($body, $offset){
 788     $me = 'sq_getnxtag';
 789     if ($offset > strlen($body)){
 790         return false;
 791     }
 792     $lt = sq_findnxstr($body, $offset, "<");
 793     if ($lt == strlen($body)){
 794         return false;
 795     }
 796     /**
 797      * We are here:
 798      * blah blah <tag attribute="value">
 799      * \---------^
 800      */
 801     $pos = sq_skipspace($body, $lt+1);
 802     if ($pos >= strlen($body)){
 803         return Array(false, false, false, $lt, strlen($body));
 804     }
 805     /**
 806      * There are 3 kinds of tags:
 807      * 1. Opening tag, e.g.:
 808      *    <a href="blah">
 809      * 2. Closing tag, e.g.:
 810      *    </a>
 811      * 3. XHTML-style content-less tag, e.g.:
 812      *    <img src="blah"/>
 813      */
 814     $tagtype = false;
 815     switch (substr($body, $pos, 1)){
 816         case '/':
 817             $tagtype = 2;
 818             $pos++;
 819             break;
 820         case '!':
 821             /**
 822              * A comment or an SGML declaration.
 823              */
 824             if (substr($body, $pos+1, 2) == "--"){
 825                 $gt = strpos($body, "-->", $pos);
 826                 if ($gt === false){
 827                     $gt = strlen($body);
 828                 } else {
 829                     $gt += 2;
 830                 }
 831                 return Array(false, false, false, $lt, $gt);
 832             } else {
 833                 $gt = sq_findnxstr($body, $pos, ">");
 834                 return Array(false, false, false, $lt, $gt);
 835             }
 836             break;
 837         default:
 838             /**
 839              * Assume tagtype 1 for now. If it's type 3, we'll switch values
 840              * later.
 841              */
 842             $tagtype = 1;
 843             break;
 844     }
 845
 846     $tag_start = $pos;
 847     $tagname = '';
 848     /**
 849      * Look for next [\W-_], which will indicate the end of the tag name.
 850      */
 851     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
 852     if ($regary == false){
 853         return Array(false, false, false, $lt, strlen($body));
 854     }
 855     list($pos, $tagname, $match) = $regary;
 856     $tagname = strtolower($tagname);
 857
 858     /**
 859      * $match can be either of these:
 860      * '>'  indicating the end of the tag entirely.
 861      * '\s' indicating the end of the tag name.
 862      * '/'  indicating that this is type-3 xhtml tag.
 863      *
 864      * Whatever else we find there indicates an invalid tag.
 865      */
 866     switch ($match){
 867         case '/':
 868             /**
 869              * This is an xhtml-style tag with a closing / at the
 870              * end, like so: <img src="blah"/>. Check if it's followed
 871              * by the closing bracket. If not, then this tag is invalid
 872              */
 873             if (substr($body, $pos, 2) == "/>"){
 874                 $pos++;
 875                 $tagtype = 3;
 876             } else {
 877                 $gt = sq_findnxstr($body, $pos, ">");
 878                 $retary = Array(false, false, false, $lt, $gt);
 879                 return $retary;
 880             }
 881         case '>':
 882             return Array($tagname, false, $tagtype, $lt, $pos);
 883             break;
 884         default:
 885             /**
 886              * Check if it's whitespace
 887              */
 888             if (!preg_match('/\s/', $match)){
 889                 /**
 890                  * This is an invalid tag! Look for the next closing ">".
 891                  */
 892                 $gt = sq_findnxstr($body, $offset, ">");
 893                 return Array(false, false, false, $lt, $gt);
 894             }
 895             break;
 896     }
 897
 898     /**
 899      * At this point we're here:
 900      * <tagname  attribute='blah'>
 901      * \-------^
 902      *
 903      * At this point we loop in order to find all attributes.
 904      */
 905     $attname = '';
 906     $atttype = false;
 907     $attary = Array();
 908
 909     while ($pos <= strlen($body)){
 910         $pos = sq_skipspace($body, $pos);
 911         if ($pos == strlen($body)){
 912             /**
 913              * Non-closed tag.
 914              */
 915             return Array(false, false, false, $lt, $pos);
 916         }
 917         /**
 918          * See if we arrived at a ">" or "/>", which means that we reached
 919          * the end of the tag.
 920          */
 921         $matches = Array();
 922         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
 923             /**
 924              * Yep. So we did.
 925              */
 926             $pos += strlen($matches{1});
 927             if ($matches{2} == "/>"){
 928                 $tagtype = 3;
 929                 $pos++;
 930             }
 931             return Array($tagname, $attary, $tagtype, $lt, $pos);
 932         }
 933
 934         /**
 935          * There are several types of attributes, with optional
 936          * [:space:] between members.
 937          * Type 1:
 938          *   attrname[:space:]=[:space:]'CDATA'
 939          * Type 2:
 940          *   attrname[:space:]=[:space:]"CDATA"
 941          * Type 3:
 942          *   attr[:space:]=[:space:]CDATA
 943          * Type 4:
 944          *   attrname
 945          *
 946          * We leave types 1 and 2 the same, type 3 we check for
 947          * '"' and convert to "&quot" if needed, then wrap in
 948          * double quotes. Type 4 we convert into:
 949          * attrname="yes".
 950          */
 951         $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
 952         if ($regary == false){
 953             /**
 954              * Looks like body ended before the end of tag.
 955              */
 956             return Array(false, false, false, $lt, strlen($body));
 957         }
 958         list($pos, $attname, $match) = $regary;
 959         $attname = strtolower($attname);
 960         /**
 961          * We arrived at the end of attribute name. Several things possible
 962          * here:
 963          * '>'  means the end of the tag and this is attribute type 4
 964          * '/'  if followed by '>' means the same thing as above
 965          * '\s' means a lot of things -- look what it's followed by.
 966          *      anything else means the attribute is invalid.
 967          */
 968         switch($match){
 969             case '/':
 970                 /**
 971                  * This is an xhtml-style tag with a closing / at the
 972                  * end, like so: <img src="blah"/>. Check if it's followed
 973                  * by the closing bracket. If not, then this tag is invalid
 974                  */
 975                 if (substr($body, $pos, 2) == "/>"){
 976                     $pos++;
 977                     $tagtype = 3;
 978                 } else {
 979                     $gt = sq_findnxstr($body, $pos, ">");
 980                     $retary = Array(false, false, false, $lt, $gt);
 981                     return $retary;
 982                 }
 983             case '>':
 984                 $attary{$attname} = '"yes"';
 985                 return Array($tagname, $attary, $tagtype, $lt, $pos);
 986                 break;
 987             default:
 988                 /**
 989                  * Skip whitespace and see what we arrive at.
 990                  */
 991                 $pos = sq_skipspace($body, $pos);
 992                 $char = substr($body, $pos, 1);
 993                 /**
 994                  * Two things are valid here:
 995                  * '=' means this is attribute type 1 2 or 3.
 996                  * \w means this was attribute type 4.
 997                  * anything else we ignore and re-loop. End of tag and
 998                  * invalid stuff will be caught by our checks at the beginning
 999                  * of the loop.
1000                  */
1001                 if ($char == "="){
1002                     $pos++;
1003                     $pos = sq_skipspace($body, $pos);
1004                     /**
1005                      * Here are 3 possibilities:
1006                      * "'"  attribute type 1
1007                      * '"'  attribute type 2
1008                      * everything else is the content of tag type 3
1009                      */
1010                     $quot = substr($body, $pos, 1);
1011                     if ($quot == "'"){
1012                         $regary = sq_findnxreg($body, $pos+1, "\'");
1013                         if ($regary == false){
1014                             return Array(false, false, false, $lt, strlen($body));
1015                         }
1016                         list($pos, $attval, $match) = $regary;
1017                         $pos++;
1018                         $attary{$attname} = "'" . $attval . "'";
1019                     } else if ($quot == '"'){
1020                         $regary = sq_findnxreg($body, $pos+1, '\"');
1021                         if ($regary == false){
1022                             return Array(false, false, false, $lt, strlen($body));
1023                         }
1024                         list($pos, $attval, $match) = $regary;
1025                         $pos++;
1026                         $attary{$attname} = '"' . $attval . '"';
1027                     } else {
1028                         /**
1029                          * These are hateful. Look for \s, or >.
1030                          */
1031                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1032                         if ($regary == false){
1033                             return Array(false, false, false, $lt, strlen($body));
1034                         }
1035                         list($pos, $attval, $match) = $regary;
1036                         /**
1037                          * If it's ">" it will be caught at the top.
1038                          */
1039                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1040                         $attary{$attname} = '"' . $attval . '"';
1041                     }
1042                 } else if (preg_match("|[\w/>]|", $char)) {
1043                     /**
1044                      * That was attribute type 4.
1045                      */
1046                     $attary{$attname} = '"yes"';
1047                 } else {
1048                     /**
1049                      * An illegal character. Find next '>' and return.
1050                      */
1051                     $gt = sq_findnxstr($body, $pos, ">");
1052                     return Array(false, false, false, $lt, $gt);
1053                 }
1054                 break;
1055         }
1056     }
1057     /**
1058      * The fact that we got here indicates that the tag end was never
1059      * found. Return invalid tag indication so it gets stripped.
1060      */
1061     return Array(false, false, false, $lt, strlen($body));
1062 }
1063
1064 /**
1065  * This function checks attribute values for entity-encoded values
1066  * and returns them translated into 8-bit strings so we can run
1067  * checks on them.
1068  *
1069  * @param  $attvalue A string to run entity check against.
1070  * @return           Translated value.
1071  */
1072 function sq_deent($attvalue){
1073     $me = 'sq_deent';
1074     /**
1075      * See if we have to run the checks first. All entities must start
1076      * with "&".
1077      */
1078     if (strpos($attvalue, "&") === false){
1079         return $attvalue;
1080     }
1081     /**
1082      * Check named entities first.
1083      */
1084     $trans = get_html_translation_table(HTML_ENTITIES);
1085     /**
1086      * Leave &quot; in, as it can mess us up.
1087      */
1088     $trans = array_flip($trans);
1089     unset($trans{"&quot;"});
1090     while (list($ent, $val) = each($trans)){
1091         $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1092     }
1093     /**
1094      * Now translate numbered entities from 1 to 255 if needed.
1095      */
1096     if (strpos($attvalue, "#") !== false){
1097         $omit = Array(34, 39);
1098         for ($asc=1; $asc<256; $asc++){
1099             if (!in_array($asc, $omit)){
1100                 $chr = chr($asc);
1101                 $attvalue = preg_replace("/\&#0*$asc;*(\D)/si", "$chr\\1",
1102                                          $attvalue);
1103                 $attvalue = preg_replace("/\&#x0*".dechex($asc).";*(\W)/si",
1104                                          "$chr\\1", $attvalue);
1105             }
1106         }
1107     }
1108     return $attvalue;
1109 }
1110
1111 /**
1112  * This function runs various checks against the attributes.
1113  *
1114  * @param  $tagname         String with the name of the tag.
1115  * @param  $attary          Array with all tag attributes.
1116  * @param  $rm_attnames     See description for sq_sanitize
1117  * @param  $bad_attvals     See description for sq_sanitize
1118  * @param  $add_attr_to_tag See description for sq_sanitize
1119  * @param  $message         message object
1120  * @param  $id              message id
1121  * @return                  Array with modified attributes.
1122  */
1123 function sq_fixatts($tagname,
1124                     $attary,
1125                     $rm_attnames,
1126                     $bad_attvals,
1127                     $add_attr_to_tag,
1128                     $message,
1129                     $id,
1130                     $mailbox
1131                     ){
1132     $me = 'sq_fixatts';
1133     while (list($attname, $attvalue) = each($attary)){
1134         /**
1135          * See if this attribute should be removed.
1136          */
1137         foreach ($rm_attnames as $matchtag=>$matchattrs){
1138             if (preg_match($matchtag, $tagname)){
1139                 foreach ($matchattrs as $matchattr){
1140                     if (preg_match($matchattr, $attname)){
1141                         unset($attary{$attname});
1142                         continue;
1143                     }
1144                 }
1145             }
1146         }
1147         /**
1148          * Remove any entities.
1149          */
1150         $attvalue = sq_deent($attvalue);
1151
1152         /**
1153          * Now let's run checks on the attvalues.
1154          * I don't expect anyone to comprehend this. If you do,
1155          * get in touch with me so I can drive to where you live and
1156          * shake your hand personally. :)
1157          */
1158         foreach ($bad_attvals as $matchtag=>$matchattrs){
1159             if (preg_match($matchtag, $tagname)){
1160                 foreach ($matchattrs as $matchattr=>$valary){
1161                     if (preg_match($matchattr, $attname)){
1162                         /**
1163                          * There are two arrays in valary.
1164                          * First is matches.
1165                          * Second one is replacements
1166                          */
1167                         list($valmatch, $valrepl) = $valary;
1168                         $newvalue =
1169                             preg_replace($valmatch, $valrepl, $attvalue);
1170                         if ($newvalue != $attvalue){
1171                             $attary{$attname} = $newvalue;
1172                         }
1173                     }
1174                 }
1175             }
1176         }
1177         /**
1178          * Turn cid: urls into http-friendly ones.
1179          */
1180         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1181             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1182         }
1183     }
1184     /**
1185      * See if we need to append any attributes to this tag.
1186      */
1187     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1188         if (preg_match($matchtag, $tagname)){
1189             $attary = array_merge($attary, $addattary);
1190         }
1191     }
1192     return $attary;
1193 }
1194
1195 /**
1196  * This function edits the style definition to make them friendly and
1197  * usable in squirrelmail.
1198  *
1199  * @param  $message  the message object
1200  * @param  $id       the message id
1201  * @param  $content  a string with whatever is between <style> and </style>
1202  * @return           a string with edited content.
1203  */
1204 function sq_fixstyle($message, $id, $content){
1205     global $view_unsafe_images;
1206     $me = 'sq_fixstyle';
1207     /**
1208      * First look for general BODY style declaration, which would be
1209      * like so:
1210      * body {background: blah-blah}
1211      * and change it to .bodyclass so we can just assign it to a <div>
1212      */
1213     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1214     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1215     /**
1216      * Fix url('blah') declarations.
1217      */
1218     $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1219                             "url(\\1$secremoveimg\\2)", $content);
1220     /**
1221      * Fix url('https*://.*) declarations but only if $view_unsafe_images
1222      * is false.
1223      */
1224     if (!$view_unsafe_images){
1225         $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1226                                 "url(\\1$secremoveimg\\2)", $content);
1227     }
1228
1229     /**
1230      * Fix urls that refer to cid:
1231      */
1232     while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1233                       $matches)){
1234         $cidurl = $matches{1};
1235         $httpurl = sq_cid2http($message, $id, $cidurl);
1236         $content = preg_replace("|url\($cidurl\)|si",
1237                                 "url($httpurl)", $content);
1238     }
1239
1240     /**
1241      * Fix stupid css declarations which lead to vulnerabilities
1242      * in IE.
1243      */
1244     $match   = Array('/expression/si',
1245                      '/behaviou*r/si',
1246                      '/binding/si');
1247     $replace = Array('idiocy', 'idiocy', 'idiocy');
1248     $content = preg_replace($match, $replace, $content);
1249     return $content;
1250 }
1251
1252 /**
1253  * This function converts cid: url's into the ones that can be viewed in
1254  * the browser.
1255  *
1256  * @param  $message  the message object
1257  * @param  $id       the message id
1258  * @param  $cidurl   the cid: url.
1259  * @return           a string with a http-friendly url
1260  */
1261 function sq_cid2http($message, $id, $cidurl, $mailbox){
1262     /**
1263      * Get rid of quotes.
1264      */
1265     $quotchar = substr($cidurl, 0, 1);
1266     $cidurl = str_replace($quotchar, "", $cidurl);
1267     $cidurl = substr(trim($cidurl), 4);
1268     $linkurl = find_ent_id($cidurl, $message);
1269     /* in case of non-save cid links $httpurl should be replaced by a sort of
1270        unsave link image */
1271     $httpurl = '';
1272     if ($linkurl) {
1273         $httpurl = $quotchar . '../src/download.php?absolute_dl=true&amp;' .
1274                    "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1275                    '&amp;ent_id=' . $linkurl . $quotchar;
1276     }
1277     return $httpurl;
1278 }
1279
1280 /**
1281  * This function changes the <body> tag into a <div> tag since we
1282  * can't really have a body-within-body.
1283  *
1284  * @param  $attary  an array of attributes and values of <body>
1285  * @return          a modified array of attributes to be set for <div>
1286  */
1287 function sq_body2div($attary){
1288     $me = 'sq_body2div';
1289     $divattary = Array('class' => "'bodyclass'");
1290     $bgcolor = '#ffffff';
1291     $text = '#000000';
1292     $styledef = '';
1293     if (is_array($attary) && sizeof($attary) > 0){
1294         foreach ($attary as $attname=>$attvalue){
1295             $quotchar = substr($attvalue, 0, 1);
1296             $attvalue = str_replace($quotchar, "", $attvalue);
1297             switch ($attname){
1298                 case 'background':
1299                     $styledef .= "background-image: url('$attvalue'); ";
1300                     break;
1301                 case 'bgcolor':
1302                     $styledef .= "background-color: $attvalue; ";
1303                     break;
1304                 case 'text':
1305                     $styledef .= "color: $attvalue; ";
1306                     break;
1307             }
1308         }
1309         if (strlen($styledef) > 0){
1310             $divattary{"style"} = "\"$styledef\"";
1311         }
1312     }
1313     return $divattary;
1314 }
1315
1316 /**
1317  * This is the main function and the one you should actually be calling.
1318  * There are several variables you should be aware of an which need
1319  * special description.
1320  *
1321  * Since the description is quite lengthy, see it here:
1322  * http://www.mricon.com/html/phpfilter.html
1323  *
1324  * @param $body                 the string with HTML you wish to filter
1325  * @param $tag_list             see description above
1326  * @param $rm_tags_with_content see description above
1327  * @param $self_closing_tags    see description above
1328  * @param $force_tag_closing    see description above
1329  * @param $rm_attnames          see description above
1330  * @param $bad_attvals          see description above
1331  * @param $add_attr_to_tag      see description above
1332  * @param $message              message object
1333  * @param $id                   message id
1334  * @return                      sanitized html safe to show on your pages.
1335  */
1336 function sq_sanitize($body,
1337                      $tag_list,
1338                      $rm_tags_with_content,
1339                      $self_closing_tags,
1340                      $force_tag_closing,
1341                      $rm_attnames,
1342                      $bad_attvals,
1343                      $add_attr_to_tag,
1344                      $message,
1345                      $id,
1346                      $mailbox
1347                      ){
1348     $me = 'sq_sanitize';
1349     /**
1350      * Normalize rm_tags and rm_tags_with_content.
1351      */
1352     @array_walk($rm_tags, 'sq_casenormalize');
1353     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1354     @array_walk($self_closing_tags, 'sq_casenormalize');
1355     /**
1356      * See if tag_list is of tags to remove or tags to allow.
1357      * false  means remove these tags
1358      * true   means allow these tags
1359      */
1360     $rm_tags = array_shift($tag_list);
1361     $curpos = 0;
1362     $open_tags = Array();
1363     $trusted = "<!-- begin sanitized html -->\n";
1364     $skip_content = false;
1365     /**
1366      * Take care of netscape's stupid javascript entities like
1367      * &{alert('boo')};
1368      */
1369     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1370
1371     while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1372         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1373         $free_content = substr($body, $curpos, $lt-$curpos);
1374         /**
1375          * Take care of <style>
1376          */
1377         if ($tagname == "style" && $tagtype == 2){
1378             /**
1379              * This is a closing </style>. Edit the
1380              * content before we apply it.
1381              */
1382             $free_content = sq_fixstyle($message, $id, $free_content);
1383         }
1384         if ($skip_content == false){
1385             $trusted .= $free_content;
1386         }
1387         if ($tagname != FALSE){
1388             if ($tagtype == 2){
1389                 if ($skip_content == $tagname){
1390                     /**
1391                      * Got to the end of tag we needed to remove.
1392                      */
1393                     $tagname = false;
1394                     $skip_content = false;
1395                 } else {
1396                     if ($skip_content == false){
1397                         if ($tagname == "body"){
1398                             $tagname = "div";
1399                         } else {
1400                             if (isset($open_tags{$tagname}) &&
1401                                 $open_tags{$tagname} > 0){
1402                                 $open_tags{$tagname}--;
1403                             } else {
1404                                 $tagname = false;
1405                             }
1406                         }
1407                     }
1408                 }
1409             } else {
1410                 /**
1411                  * $rm_tags_with_content
1412                  */
1413                 if ($skip_content == false){
1414                     /**
1415                      * See if this is a self-closing type and change
1416                      * tagtype appropriately.
1417                      */
1418                     if ($tagtype == 1
1419                         && in_array($tagname, $self_closing_tags)){
1420                         $tagtype=3;
1421                     }
1422                     /**
1423                      * See if we should skip this tag and any content
1424                      * inside it.
1425                      */
1426                     if ($tagtype == 1 &&
1427                         in_array($tagname, $rm_tags_with_content)){
1428                         $skip_content = $tagname;
1429                     } else {
1430                         if (($rm_tags == false
1431                              && in_array($tagname, $tag_list)) ||
1432                             ($rm_tags == true &&
1433                              !in_array($tagname, $tag_list))){
1434                             $tagname = false;
1435                         } else {
1436                             if ($tagtype == 1){
1437                                 if (isset($open_tags{$tagname})){
1438                                     $open_tags{$tagname}++;
1439                                 } else {
1440                                     $open_tags{$tagname}=1;
1441                                 }
1442                             }
1443                             /**
1444                              * This is where we run other checks.
1445                              */
1446                             if (is_array($attary) && sizeof($attary) > 0){
1447                                 $attary = sq_fixatts($tagname,
1448                                                      $attary,
1449                                                      $rm_attnames,
1450                                                      $bad_attvals,
1451                                                      $add_attr_to_tag,
1452                                                      $message,
1453                                                      $id,
1454                                                      $mailbox
1455                                                      );
1456                             }
1457                             /**
1458                              * Convert body into div.
1459                              */
1460                             if ($tagname == "body"){
1461                                 $tagname = "div";
1462                                 $attary = sq_body2div($attary, $message, $id);
1463                             }
1464                         }
1465                     }
1466                 }
1467             }
1468             if ($tagname != false && $skip_content == false){
1469                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1470             }
1471         }
1472         $curpos = $gt+1;
1473     }
1474     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1475     if ($force_tag_closing == true){
1476         foreach ($open_tags as $tagname=>$opentimes){
1477             while ($opentimes > 0){
1478                 $trusted .= '</' . $tagname . '>';
1479                 $opentimes--;
1480             }
1481         }
1482         $trusted .= "\n";
1483     }
1484     $trusted .= "<!-- end sanitized html -->\n";
1485     return $trusted;
1486 }
1487
1488 /**
1489  * This is a wrapper function to call html sanitizing routines.
1490  *
1491  * @param  $body  the body of the message
1492  * @param  $id    the id of the message
1493  * @return        a string with html safe to display in the browser.
1494  */
1495 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1496     global $attachment_common_show_images, $view_unsafe_images,
1497            $has_unsafe_images;
1498     /**
1499      * Don't display attached images in HTML mode.
1500      */
1501     $attachment_common_show_images = false;
1502     $tag_list = Array(
1503                       false,
1504                       "object",
1505                       "meta",
1506                       "html",
1507                       "head",
1508                       "base",
1509                       "link",
1510                       "frame",
1511                       "iframe"
1512                       );
1513
1514     $rm_tags_with_content = Array(
1515                                   "script",
1516                                   "applet",
1517                                   "embed",
1518                                   "title"
1519                                   );
1520
1521     $self_closing_tags =  Array(
1522                                 "img",
1523                                 "br",
1524                                 "hr",
1525                                 "input"
1526                                 );
1527
1528     $force_tag_closing = false;
1529
1530     $rm_attnames = Array(
1531                          "/.*/" =>
1532                          Array(
1533                                "/target/si",
1534                                "/^on.*/si",
1535                                "/^dynsrc/si",
1536                                "/^data.*/si"
1537                                )
1538                          );
1539
1540     $secremoveimg = "../images/" . _("sec_remove_eng.png");
1541     $bad_attvals = Array(
1542         "/.*/" =>
1543             Array(
1544                 "/^src|background/i" =>
1545                     Array(
1546                           Array(
1547                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1548                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1549                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1550                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1551                                 ),
1552                           Array(
1553                                 "\\1$secremoveimg\\2",
1554                                 "\\1$secremoveimg\\2",
1555                                 "\\1$secremoveimg\\2",
1556                                 "\\1$secremoveimg\\2"
1557                                 )
1558                         ),
1559                 "/^href|action/i" =>
1560                     Array(
1561                           Array(
1562                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1563                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1564                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1565                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1566                                 ),
1567                           Array(
1568                                 "\\1#\\2",
1569                                 "\\1#\\2",
1570                                 "\\1#\\2",
1571                                 "\\1#\\2"
1572                                 )
1573                         ),
1574                 "/^style/si" =>
1575                     Array(
1576                           Array(
1577                                 "/expression/si",
1578                                 "/binding/si",
1579                                 "/behaviou*r/si",
1580                                 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1581                                 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1582                                 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1583                                 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1584                                ),
1585                           Array(
1586                                 "idiocy",
1587                                 "idiocy",
1588                                 "idiocy",
1589                                 "url(\\1#\\2)",
1590                                 "url(\\1#\\2)",
1591                                 "url(\\1#\\2)",
1592                                 "url(\\1#\\2)"
1593                                )
1594                           )
1595                 )
1596         );
1597     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1598         $view_unsafe_images = false;
1599     }
1600     if (!$view_unsafe_images){
1601         /**
1602          * Remove any references to http/https if view_unsafe_images set
1603          * to false.
1604          */
1605          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1606                     '/^([\'\"])\s*https*:.*([\'\"])/si');
1607          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1608                     "\\1$secremoveimg\\2");
1609          array_push($bad_attvals{'/.*/'}{'/^style/si'}[0],
1610                     '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1611          array_push($bad_attvals{'/.*/'}{'/^style/si'}[1],
1612                     "url(\\1$secremoveimg\\2)");
1613     }
1614
1615     $add_attr_to_tag = Array(
1616                              "/^a$/si" => Array('target'=>'"_new"')
1617                              );
1618     $trusted = sq_sanitize($body,
1619                            $tag_list,
1620                            $rm_tags_with_content,
1621                            $self_closing_tags,
1622                            $force_tag_closing,
1623                            $rm_attnames,
1624                            $bad_attvals,
1625                            $add_attr_to_tag,
1626                            $message,
1627                            $id,
1628                            $mailbox
1629                            );
1630     if (preg_match("|$secremoveimg|si", $trusted)){
1631         $has_unsafe_images = true;
1632     }
1633     return $trusted;
1634 }
1635
1636 ?>