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