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 % 2) {
 294             if (!isset($color[13])) {
 295                 $color[13] = '#800000';
 296             }
 297             $line = '<font color="' . $color[13] . '">' . $line . '</font>';
 298         } elseif ($quotes) {
 299             if (!isset($color[14])) {
 300                 $color[14] = '#FF0000';
 301             }
 302             $line = '<font color="' . $color[14] . '">' . $line . '</font>';
 303         }
 304
 305         $body_ary[$i] = $line;
 306     }
 307     $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
 308 }
 309
 310 /* This returns a parsed string called $body. That string can then
 311  * be displayed as the actual message in the HTML. It contains
 312  * everything needed, including HTML Tags, Attachments at the
 313  * bottom, etc.
 314  */
 315 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
 316     /* This if statement checks for the entity to show as the
 317      * primary message. To add more of them, just put them in the
 318      * order that is their priority.
 319      */
 320     global $startMessage, $username, $key, $imapServerAddress, $imapPort,
 321            $show_html_default, $sort, $has_unsafe_images, $passed_ent_id;
 322
 323     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
 324         $view_unsafe_images = false;
 325     }
 326
 327     $body = '';
 328     $urlmailbox = urlencode($mailbox);
 329     $body_message = getEntity($message, $ent_num);
 330     if (($body_message->header->type0 == 'text') ||
 331         ($body_message->header->type0 == 'rfc822')) {
 332         $body = mime_fetch_body ($imap_stream, $id, $ent_num);
 333         $body = decodeBody($body, $body_message->header->encoding);
 334         $hookResults = do_hook("message_body", $body);
 335         $body = $hookResults[1];
 336
 337         /* If there are other types that shouldn't be formatted, add
 338          * them here.
 339          */
 340
 341         if ($body_message->header->type1 == 'html') {
 342             if ($show_html_default <> 1) {
 343                 $entity_conv = array('&nbsp;' => ' ',
 344                                      '<p>'    => "\n",
 345                                      '<br>'   => "\n",
 346                                      '<P>'    => "\n",
 347                                      '<BR>'   => "\n",
 348                                      '&gt;'   => '>',
 349                                      '&lt;'   => '<');
 350                 $body = strtr($body, $entity_conv);
 351                 $body = strip_tags($body);
 352                 $body = trim($body);
 353                 translateText($body, $wrap_at,
 354                               $body_message->header->getParameter('charset'));
 355             } else {
 356                 $body = magicHTML($body, $id, $message, $mailbox);
 357             }
 358         } else {
 359             translateText($body, $wrap_at,
 360                           $body_message->header->getParameter('charset'));
 361         }
 362         $link = 'read_body.php?passed_id=' . $id . '&amp;ent_id='.$ent_num.
 363                  '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
 364                  '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 365         if (isset($passed_ent_id)) {
 366             $link .= '&amp;passed_ent_id='.$passed_ent_id;
 367         }
 368         if ($view_unsafe_images) {
 369             $text = _("Hide Unsafe Images");
 370         } else {
 371             if (isset($has_unsafe_images) && $has_unsafe_images) {
 372                 $link .= '&amp;view_unsafe_images=1';
 373                 $text = _("View Unsafe Images");
 374             } else {
 375                 $text = '';
 376             }
 377         }
 378         $body .= '<center><small><a href="'.$link.'">'.$text.
 379                  '</a></small></center><br>' . "\n";
 380     }
 381     return $body;
 382 }
 383
 384
 385 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 386     global $where, $what, $startMessage, $color;
 387     static $ShownHTML = 0;
 388
 389     $att_ar = $message->getAttachments($exclude_id);
 390
 391     if (!count($att_ar)) return '';
 392
 393     $attachments = '';
 394
 395     $urlMailbox = urlencode($mailbox);
 396
 397     foreach ($att_ar as $att) {
 398         $ent = urldecode($att->entity_id);
 399         $header = $att->header;
 400         $type0 = strtolower($header->type0);
 401         $type1 = strtolower($header->type1);
 402         $name = '';
 403         $links['download link']['text'] = _("download");
 404         $links['download link']['href'] =
 405                 "../src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 406         $ImageURL = '';
 407         if ($type0 =='message' && $type1 == 'rfc822') {
 408             $default_page = '../src/read_body.php';
 409             $rfc822_header = $att->rfc822_header;
 410             $filename = $rfc822_header->subject;
 411             if (trim( $filename ) == '') {
 412                 $filename = 'untitled-[' . $ent . ']' ;
 413             }
 414             $from_o = $rfc822_header->from;
 415             if (is_object($from_o)) {
 416                 $from_name = $from_o->getAddress(false);
 417             } else {
 418                 $from_name = _("Unknown sender");
 419             }
 420             $from_name = decodeHeader(($from_name));
 421             $description = $from_name;
 422         } else {
 423             $default_page = '../src/download.php';
 424             if (is_object($header->disposition)) {
 425                 $filename = $header->disposition->getProperty('filename');
 426                 if (trim($filename) == '') {
 427                     $name = decodeHeader($header->disposition->getProperty('name'));
 428                     if (trim($name) == '') {
 429                         $name = $header->getParameter('name');
 430                         if(trim($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                     } else {
 440                         $filename = $name;
 441                     }
 442                 }
 443             } else {
 444                 $filename = $header->getParameter('name');
 445                 if (!trim($filename)) {
 446                     if (trim( $header->id ) == '') {
 447                         $filename = 'untitled-[' . $ent . ']' ;
 448                     } else {
 449                         $filename = 'cid: ' . $header->id;
 450                     }
 451                 }
 452             }
 453             if ($header->description) {
 454                 $description = decodeHeader($header->description);
 455             } else {
 456                 $description = '';
 457             }
 458         }
 459
 460         $display_filename = $filename;
 461         if (isset($passed_ent_id)) {
 462             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 463         } else {
 464             $passed_ent_id_link = '';
 465         }
 466         $defaultlink = $default_page . "?startMessage=$startMessage"
 467                      . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 468                      . '&amp;ent_id='.$ent.$passed_ent_id_link.'&amp;absolute_dl=true';
 469         if ($where && $what) {
 470            $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 471         }
 472         /* This executes the attachment hook with a specific MIME-type.
 473          * If that doesn't have results, it tries if there's a rule
 474          * for a more generic type.
 475          */
 476         $hookresults = do_hook("attachment $type0/$type1", $links,
 477                                $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 478                                $display_filename, $where, $what);
 479         if(count($hookresults[1]) <= 1) {
 480             $hookresults = do_hook("attachment $type0/*", $links,
 481                                    $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 482                                    $display_filename, $where, $what);
 483         }
 484
 485         $links = $hookresults[1];
 486         $defaultlink = $hookresults[6];
 487
 488         $attachments .= '<TR><TD>' .
 489                         '<A HREF="'.$defaultlink.'">'.decodeHeader($display_filename).'</A>&nbsp;</TD>' .
 490                         '<TD><SMALL><b>' . show_readable_size($header->size) .
 491                         '</b>&nbsp;&nbsp;</small></TD>' .
 492                         "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
 493                         '<TD><SMALL>';
 494         $attachments .= '<b>' . $description . '</b>';
 495         $attachments .= '</SMALL></TD><TD><SMALL>&nbsp;';
 496
 497         $skipspaces = 1;
 498         foreach ($links as $val) {
 499             if ($skipspaces) {
 500                 $skipspaces = 0;
 501             } else {
 502                 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
 503             }
 504             $attachments .= '<a href="' . $val['href'] . '">' .  $val['text'] . '</a>';
 505         }
 506         unset($links);
 507         $attachments .= "</TD></TR>\n";
 508     }
 509     return $attachments;
 510 }
 511
 512 /* This function decodes the body depending on the encoding type. */
 513 function decodeBody($body, $encoding) {
 514     global $languages, $squirrelmail_language;
 515     global $show_html_default;
 516
 517     $body = str_replace("\r\n", "\n", $body);
 518     $encoding = strtolower($encoding);
 519
 520     if ($encoding == 'quoted-printable' ||
 521         $encoding == 'quoted_printable') {
 522         $body = quoted_printable_decode($body);
 523
 524         while (ereg("=\n", $body)) {
 525             $body = ereg_replace ("=\n", '', $body);
 526         }
 527
 528     } else if ($encoding == 'base64') {
 529         $body = base64_decode($body);
 530     }
 531
 532     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 533         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 534         $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
 535     }
 536
 537     // All other encodings are returned raw.
 538     return $body;
 539 }
 540
 541 /*
 542  * This functions decode strings that is encoded according to
 543  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 544  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 545  */
 546 function decodeHeader ($string, $utfencode=true,$htmlsave=true) {
 547     global $languages, $squirrelmail_language;
 548     if (is_array($string)) {
 549         $string = implode("\n", $string);
 550     }
 551
 552     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 553         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 554         $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
 555     }
 556
 557     $i = 0;
 558     $aString = explode(' ',$string);
 559     foreach ($aString as $chunk) {
 560         $encoded = false;
 561         $aString[$i] = '';
 562         while (preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 563             $aString[$i] .= $res[1];
 564             $encoding = ucfirst($res[3]);
 565             switch ($encoding)
 566             {
 567             case 'B':
 568                 $replace = base64_decode($res[4]);
 569                 $aString[$i] .= charset_decode($res[2],$replace);
 570                 break;
 571             case 'Q':
 572
 573                 $replace = str_replace('_', ' ', $res[4]);
 574                 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 575                                     $replace);
 576                 /* Only encode into entities by default. Some places
 577                  * don't need the encoding, like the compose form.
 578                  */
 579                 if ($utfencode) {
 580                     $replace = charset_decode($res[2], $replace);
 581                 } else {
 582                     if ($htmlsave) {
 583                         $replace = htmlspecialchars($res[4]);
 584                     }
 585                 }
 586                 $aString[$i] .= $replace;
 587                 break;
 588             default:
 589                 break;
 590             }
 591             $chunk = $res[5];
 592             $encoded = true;
 593         }
 594         if (!$encoded && $htmlsave) {
 595             $aString[$i] = htmlspecialchars($chunk);
 596         } else {
 597             $aString[$i] .= $chunk;
 598         }
 599         ++$i;
 600     }
 601     if (!$htmlsave) {
 602         return implode(' ',$aString);
 603     } else {
 604         return implode('&nbsp;',$aString);
 605     }
 606 }
 607
 608 /*
 609  * Encode a string according to RFC 1522 for use in headers if it
 610  * contains 8-bit characters or anything that looks like it should
 611  * be encoded.
 612  */
 613 function encodeHeader ($string) {
 614     global $default_charset, $languages, $squirrelmail_language;
 615
 616     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 617         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 618         return  $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
 619     }
 620
 621     // Encode only if the string contains 8-bit characters or =?
 622     $j = strlen($string);
 623     $l = strstr($string, '=?');         // Must be encoded ?
 624     $max_l = 75 - strlen($default_charset) - 7;
 625     $aRet = array();
 626     $ret = '';
 627     $cur_l = 0;
 628     for($i = 0; $i < $j; ++$i) {
 629         switch($string{$i}) {
 630             case '=':
 631                 $cur_l+=3;
 632                 if ($cur_l > $max_l) {
 633                     $aRet[] = "=?$default_charset?Q?$ret?=";
 634                     $cur_l = 3;
 635                     $ret = '';
 636                 }
 637                 $ret .= '=3D';
 638                 break;
 639             case '?':
 640                 $cur_l+=3;
 641                 if ($cur_l > $max_l) {
 642                     $aRet[] = "=?$default_charset?Q?$ret?=";
 643                     $cur_l = 3;
 644                     $ret = '';
 645                 }
 646                 $ret .= '=3F';
 647                 break;
 648             case '_':
 649                 $cur_l+=3;
 650                 if ($cur_l > $max_l) {
 651                     $aRet[] = "=?$default_charset?Q?$ret?=";
 652                     $cur_l = 3;
 653                     $ret = '';
 654                 }
 655                 $ret .= '=5F';
 656                 break;
 657             case ' ':
 658                 $cur_l++;
 659                 if ($cur_l > $max_l) {
 660                     $aRet[] = "=?$default_charset?Q?$ret?=";
 661                     $cur_l = 1;
 662                     $ret = '';
 663                 }
 664                 $ret .= '_';
 665                 break;
 666             default:
 667                 $k = ord($string{$i});
 668                 if ($k > 126) {
 669                     $s = sprintf("=%02X", $k);
 670                     $cur_l += strlen($s);
 671                     if ($cur_l > $max_l) {
 672                         $aRet[] = "=?$default_charset?Q?$ret?=";
 673                         $cur_l = strlen($s);
 674                         $ret = '';
 675                     }
 676                     $ret .= $s;
 677                     $l = TRUE;
 678                 } else {
 679                     $cur_l++;
 680                     if ($cur_l > $max_l) {
 681                         $aRet[] = "=?$default_charset?Q?$ret?=";
 682                         $cur_l = 1;
 683                         $ret = '';
 684                     }
 685                     $ret .= $string{$i};
 686                 }
 687                 break;
 688         }
 689     }
 690
 691     if ($l) {
 692         $string = implode('',$aRet) . "=?$default_charset?Q?$ret?=";
 693     }
 694
 695     return $string;
 696 }
 697
 698 /* This function trys to locate the entity_id of a specific mime element */
 699 function find_ent_id($id, $message) {
 700     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
 701         if ($message->entities[$i]->header->type0 == 'multipart')  {
 702             $ret = find_ent_id($id, $message->entities[$i]);
 703         } else {
 704             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
 705 //                if (sq_check_save_extension($message->entities[$i])) {
 706                     return $message->entities[$i]->entity_id;
 707 //                }
 708             }
 709         }
 710     }
 711     return $ret;
 712 }
 713
 714 function sq_check_save_extension($message) {
 715     $filename = $message->getFilename();
 716     $ext = substr($filename, strrpos($filename,'.')+1);
 717     $save_extensions = array('jpg','jpeg','gif','png','bmp');
 718     return in_array($ext, $save_extensions);
 719 }
 720
 721
 722 /**
 723  ** HTMLFILTER ROUTINES
 724  */
 725
 726 /**
 727  * This function returns the final tag out of the tag name, an array
 728  * of attributes, and the type of the tag. This function is called by
 729  * sq_sanitize internally.
 730  *
 731  * @param  $tagname  the name of the tag.
 732  * @param  $attary   the array of attributes and their values
 733  * @param  $tagtype  The type of the tag (see in comments).
 734  * @return           a string with the final tag representation.
 735  */
 736 function sq_tagprint($tagname, $attary, $tagtype){
 737     $me = 'sq_tagprint';
 738
 739     if ($tagtype == 2){
 740         $fulltag = '</' . $tagname . '>';
 741     } else {
 742         $fulltag = '<' . $tagname;
 743         if (is_array($attary) && sizeof($attary)){
 744             $atts = Array();
 745             while (list($attname, $attvalue) = each($attary)){
 746                 array_push($atts, "$attname=$attvalue");
 747             }
 748             $fulltag .= ' ' . join(" ", $atts);
 749         }
 750         if ($tagtype == 3){
 751             $fulltag .= ' /';
 752         }
 753         $fulltag .= '>';
 754     }
 755     return $fulltag;
 756 }
 757
 758 /**
 759  * A small helper function to use with array_walk. Modifies a by-ref
 760  * value and makes it lowercase.
 761  *
 762  * @param  $val a value passed by-ref.
 763  * @return      void since it modifies a by-ref value.
 764  */
 765 function sq_casenormalize(&$val){
 766     $val = strtolower($val);
 767 }
 768
 769 /**
 770  * This function skips any whitespace from the current position within
 771  * a string and to the next non-whitespace value.
 772  *
 773  * @param  $body   the string
 774  * @param  $offset the offset within the string where we should start
 775  *                 looking for the next non-whitespace character.
 776  * @return         the location within the $body where the next
 777  *                 non-whitespace char is located.
 778  */
 779 function sq_skipspace($body, $offset){
 780     $me = 'sq_skipspace';
 781     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
 782     if (sizeof($matches{1})){
 783         $count = strlen($matches{1});
 784         $offset += $count;
 785     }
 786     return $offset;
 787 }
 788
 789 /**
 790  * This function looks for the next character within a string.  It's
 791  * really just a glorified "strpos", except it catches if failures
 792  * nicely.
 793  *
 794  * @param  $body   The string to look for needle in.
 795  * @param  $offset Start looking from this position.
 796  * @param  $needle The character/string to look for.
 797  * @return         location of the next occurance of the needle, or
 798  *                 strlen($body) if needle wasn't found.
 799  */
 800 function sq_findnxstr($body, $offset, $needle){
 801     $me  = 'sq_findnxstr';
 802     $pos = strpos($body, $needle, $offset);
 803     if ($pos === FALSE){
 804         $pos = strlen($body);
 805     }
 806     return $pos;
 807 }
 808
 809 /**
 810  * This function takes a PCRE-style regexp and tries to match it
 811  * within the string.
 812  *
 813  * @param  $body   The string to look for needle in.
 814  * @param  $offset Start looking from here.
 815  * @param  $reg    A PCRE-style regex to match.
 816  * @return         Returns a false if no matches found, or an array
 817  *                 with the following members:
 818  *                 - integer with the location of the match within $body
 819  *                 - string with whatever content between offset and the match
 820  *                 - string with whatever it is we matched
 821  */
 822 function sq_findnxreg($body, $offset, $reg){
 823     $me = 'sq_findnxreg';
 824     $matches = Array();
 825     $retarr = Array();
 826     preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
 827     if (!$matches{0}){
 828         $retarr = false;
 829     } else {
 830         $retarr{0} = $offset + strlen($matches{1});
 831         $retarr{1} = $matches{1};
 832         $retarr{2} = $matches{2};
 833     }
 834     return $retarr;
 835 }
 836
 837 /**
 838  * This function looks for the next tag.
 839  *
 840  * @param  $body   String where to look for the next tag.
 841  * @param  $offset Start looking from here.
 842  * @return         false if no more tags exist in the body, or
 843  *                 an array with the following members:
 844  *                 - string with the name of the tag
 845  *                 - array with attributes and their values
 846  *                 - integer with tag type (1, 2, or 3)
 847  *                 - integer where the tag starts (starting "<")
 848  *                 - integer where the tag ends (ending ">")
 849  *                 first three members will be false, if the tag is invalid.
 850  */
 851 function sq_getnxtag($body, $offset){
 852     $me = 'sq_getnxtag';
 853     if ($offset > strlen($body)){
 854         return false;
 855     }
 856     $lt = sq_findnxstr($body, $offset, "<");
 857     if ($lt == strlen($body)){
 858         return false;
 859     }
 860     /**
 861      * We are here:
 862      * blah blah <tag attribute="value">
 863      * \---------^
 864      */
 865     $pos = sq_skipspace($body, $lt+1);
 866     if ($pos >= strlen($body)){
 867         return Array(false, false, false, $lt, strlen($body));
 868     }
 869     /**
 870      * There are 3 kinds of tags:
 871      * 1. Opening tag, e.g.:
 872      *    <a href="blah">
 873      * 2. Closing tag, e.g.:
 874      *    </a>
 875      * 3. XHTML-style content-less tag, e.g.:
 876      *    <img src="blah"/>
 877      */
 878     $tagtype = false;
 879     switch (substr($body, $pos, 1)){
 880         case '/':
 881             $tagtype = 2;
 882             $pos++;
 883             break;
 884         case '!':
 885             /**
 886              * A comment or an SGML declaration.
 887              */
 888             if (substr($body, $pos+1, 2) == "--"){
 889                 $gt = strpos($body, "-->", $pos);
 890                 if ($gt === false){
 891                     $gt = strlen($body);
 892                 } else {
 893                     $gt += 2;
 894                 }
 895                 return Array(false, false, false, $lt, $gt);
 896             } else {
 897                 $gt = sq_findnxstr($body, $pos, ">");
 898                 return Array(false, false, false, $lt, $gt);
 899             }
 900             break;
 901         default:
 902             /**
 903              * Assume tagtype 1 for now. If it's type 3, we'll switch values
 904              * later.
 905              */
 906             $tagtype = 1;
 907             break;
 908     }
 909
 910     $tag_start = $pos;
 911     $tagname = '';
 912     /**
 913      * Look for next [\W-_], which will indicate the end of the tag name.
 914      */
 915     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
 916     if ($regary == false){
 917         return Array(false, false, false, $lt, strlen($body));
 918     }
 919     list($pos, $tagname, $match) = $regary;
 920     $tagname = strtolower($tagname);
 921
 922     /**
 923      * $match can be either of these:
 924      * '>'  indicating the end of the tag entirely.
 925      * '\s' indicating the end of the tag name.
 926      * '/'  indicating that this is type-3 xhtml tag.
 927      *
 928      * Whatever else we find there indicates an invalid tag.
 929      */
 930     switch ($match){
 931         case '/':
 932             /**
 933              * This is an xhtml-style tag with a closing / at the
 934              * end, like so: <img src="blah"/>. Check if it's followed
 935              * by the closing bracket. If not, then this tag is invalid
 936              */
 937             if (substr($body, $pos, 2) == "/>"){
 938                 $pos++;
 939                 $tagtype = 3;
 940             } else {
 941                 $gt = sq_findnxstr($body, $pos, ">");
 942                 $retary = Array(false, false, false, $lt, $gt);
 943                 return $retary;
 944             }
 945         case '>':
 946             return Array($tagname, false, $tagtype, $lt, $pos);
 947             break;
 948         default:
 949             /**
 950              * Check if it's whitespace
 951              */
 952             if (!preg_match('/\s/', $match)){
 953                 /**
 954                  * This is an invalid tag! Look for the next closing ">".
 955                  */
 956                 $gt = sq_findnxstr($body, $offset, ">");
 957                 return Array(false, false, false, $lt, $gt);
 958             }
 959             break;
 960     }
 961
 962     /**
 963      * At this point we're here:
 964      * <tagname  attribute='blah'>
 965      * \-------^
 966      *
 967      * At this point we loop in order to find all attributes.
 968      */
 969     $attname = '';
 970     $atttype = false;
 971     $attary = Array();
 972
 973     while ($pos <= strlen($body)){
 974         $pos = sq_skipspace($body, $pos);
 975         if ($pos == strlen($body)){
 976             /**
 977              * Non-closed tag.
 978              */
 979             return Array(false, false, false, $lt, $pos);
 980         }
 981         /**
 982          * See if we arrived at a ">" or "/>", which means that we reached
 983          * the end of the tag.
 984          */
 985         $matches = Array();
 986         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
 987             /**
 988              * Yep. So we did.
 989              */
 990             $pos += strlen($matches{1});
 991             if ($matches{2} == "/>"){
 992                 $tagtype = 3;
 993                 $pos++;
 994             }
 995             return Array($tagname, $attary, $tagtype, $lt, $pos);
 996         }
 997
 998         /**
 999          * There are several types of attributes, with optional
1000          * [:space:] between members.
1001          * Type 1:
1002          *   attrname[:space:]=[:space:]'CDATA'
1003          * Type 2:
1004          *   attrname[:space:]=[:space:]"CDATA"
1005          * Type 3:
1006          *   attr[:space:]=[:space:]CDATA
1007          * Type 4:
1008          *   attrname
1009          *
1010          * We leave types 1 and 2 the same, type 3 we check for
1011          * '"' and convert to "&quot" if needed, then wrap in
1012          * double quotes. Type 4 we convert into:
1013          * attrname="yes".
1014          */
1015         $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1016         if ($regary == false){
1017             /**
1018              * Looks like body ended before the end of tag.
1019              */
1020             return Array(false, false, false, $lt, strlen($body));
1021         }
1022         list($pos, $attname, $match) = $regary;
1023         $attname = strtolower($attname);
1024         /**
1025          * We arrived at the end of attribute name. Several things possible
1026          * here:
1027          * '>'  means the end of the tag and this is attribute type 4
1028          * '/'  if followed by '>' means the same thing as above
1029          * '\s' means a lot of things -- look what it's followed by.
1030          *      anything else means the attribute is invalid.
1031          */
1032         switch($match){
1033             case '/':
1034                 /**
1035                  * This is an xhtml-style tag with a closing / at the
1036                  * end, like so: <img src="blah"/>. Check if it's followed
1037                  * by the closing bracket. If not, then this tag is invalid
1038                  */
1039                 if (substr($body, $pos, 2) == "/>"){
1040                     $pos++;
1041                     $tagtype = 3;
1042                 } else {
1043                     $gt = sq_findnxstr($body, $pos, ">");
1044                     $retary = Array(false, false, false, $lt, $gt);
1045                     return $retary;
1046                 }
1047             case '>':
1048                 $attary{$attname} = '"yes"';
1049                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1050                 break;
1051             default:
1052                 /**
1053                  * Skip whitespace and see what we arrive at.
1054                  */
1055                 $pos = sq_skipspace($body, $pos);
1056                 $char = substr($body, $pos, 1);
1057                 /**
1058                  * Two things are valid here:
1059                  * '=' means this is attribute type 1 2 or 3.
1060                  * \w means this was attribute type 4.
1061                  * anything else we ignore and re-loop. End of tag and
1062                  * invalid stuff will be caught by our checks at the beginning
1063                  * of the loop.
1064                  */
1065                 if ($char == "="){
1066                     $pos++;
1067                     $pos = sq_skipspace($body, $pos);
1068                     /**
1069                      * Here are 3 possibilities:
1070                      * "'"  attribute type 1
1071                      * '"'  attribute type 2
1072                      * everything else is the content of tag type 3
1073                      */
1074                     $quot = substr($body, $pos, 1);
1075                     if ($quot == "'"){
1076                         $regary = sq_findnxreg($body, $pos+1, "\'");
1077                         if ($regary == false){
1078                             return Array(false, false, false, $lt, strlen($body));
1079                         }
1080                         list($pos, $attval, $match) = $regary;
1081                         $pos++;
1082                         $attary{$attname} = "'" . $attval . "'";
1083                     } else if ($quot == '"'){
1084                         $regary = sq_findnxreg($body, $pos+1, '\"');
1085                         if ($regary == false){
1086                             return Array(false, false, false, $lt, strlen($body));
1087                         }
1088                         list($pos, $attval, $match) = $regary;
1089                         $pos++;
1090                         $attary{$attname} = '"' . $attval . '"';
1091                     } else {
1092                         /**
1093                          * These are hateful. Look for \s, or >.
1094                          */
1095                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1096                         if ($regary == false){
1097                             return Array(false, false, false, $lt, strlen($body));
1098                         }
1099                         list($pos, $attval, $match) = $regary;
1100                         /**
1101                          * If it's ">" it will be caught at the top.
1102                          */
1103                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1104                         $attary{$attname} = '"' . $attval . '"';
1105                     }
1106                 } else if (preg_match("|[\w/>]|", $char)) {
1107                     /**
1108                      * That was attribute type 4.
1109                      */
1110                     $attary{$attname} = '"yes"';
1111                 } else {
1112                     /**
1113                      * An illegal character. Find next '>' and return.
1114                      */
1115                     $gt = sq_findnxstr($body, $pos, ">");
1116                     return Array(false, false, false, $lt, $gt);
1117                 }
1118                 break;
1119         }
1120     }
1121     /**
1122      * The fact that we got here indicates that the tag end was never
1123      * found. Return invalid tag indication so it gets stripped.
1124      */
1125     return Array(false, false, false, $lt, strlen($body));
1126 }
1127
1128 /**
1129  * This function checks attribute values for entity-encoded values
1130  * and returns them translated into 8-bit strings so we can run
1131  * checks on them.
1132  *
1133  * @param  $attvalue A string to run entity check against.
1134  * @return           Translated value.
1135  */
1136 function sq_deent($attvalue){
1137     $me = 'sq_deent';
1138     /**
1139      * See if we have to run the checks first. All entities must start
1140      * with "&".
1141      */
1142     if (strpos($attvalue, "&") === false){
1143         return $attvalue;
1144     }
1145     /**
1146      * Check named entities first.
1147      */
1148     $trans = get_html_translation_table(HTML_ENTITIES);
1149     /**
1150      * Leave &quot; in, as it can mess us up.
1151      */
1152     $trans = array_flip($trans);
1153     unset($trans{"&quot;"});
1154     while (list($ent, $val) = each($trans)){
1155         $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1156     }
1157     /**
1158      * Now translate numbered entities from 1 to 255 if needed.
1159      */
1160     if (strpos($attvalue, "#") !== false){
1161         $omit = Array(34, 39);
1162         for ($asc=1; $asc<256; $asc++){
1163             if (!in_array($asc, $omit)){
1164                 $chr = chr($asc);
1165                 $attvalue = preg_replace("/\&#0*$asc;*(\D)/si", "$chr\\1",
1166                                          $attvalue);
1167                 $attvalue = preg_replace("/\&#x0*".dechex($asc).";*(\W)/si",
1168                                          "$chr\\1", $attvalue);
1169             }
1170         }
1171     }
1172     return $attvalue;
1173 }
1174
1175 /**
1176  * This function runs various checks against the attributes.
1177  *
1178  * @param  $tagname         String with the name of the tag.
1179  * @param  $attary          Array with all tag attributes.
1180  * @param  $rm_attnames     See description for sq_sanitize
1181  * @param  $bad_attvals     See description for sq_sanitize
1182  * @param  $add_attr_to_tag See description for sq_sanitize
1183  * @param  $message         message object
1184  * @param  $id              message id
1185  * @return                  Array with modified attributes.
1186  */
1187 function sq_fixatts($tagname,
1188                     $attary,
1189                     $rm_attnames,
1190                     $bad_attvals,
1191                     $add_attr_to_tag,
1192                     $message,
1193                     $id,
1194                     $mailbox
1195                     ){
1196     $me = 'sq_fixatts';
1197     while (list($attname, $attvalue) = each($attary)){
1198         /**
1199          * See if this attribute should be removed.
1200          */
1201         foreach ($rm_attnames as $matchtag=>$matchattrs){
1202             if (preg_match($matchtag, $tagname)){
1203                 foreach ($matchattrs as $matchattr){
1204                     if (preg_match($matchattr, $attname)){
1205                         unset($attary{$attname});
1206                         continue;
1207                     }
1208                 }
1209             }
1210         }
1211         /**
1212          * Remove any entities.
1213          */
1214         $attvalue = sq_deent($attvalue);
1215
1216         /**
1217          * Now let's run checks on the attvalues.
1218          * I don't expect anyone to comprehend this. If you do,
1219          * get in touch with me so I can drive to where you live and
1220          * shake your hand personally. :)
1221          */
1222         foreach ($bad_attvals as $matchtag=>$matchattrs){
1223             if (preg_match($matchtag, $tagname)){
1224                 foreach ($matchattrs as $matchattr=>$valary){
1225                     if (preg_match($matchattr, $attname)){
1226                         /**
1227                          * There are two arrays in valary.
1228                          * First is matches.
1229                          * Second one is replacements
1230                          */
1231                         list($valmatch, $valrepl) = $valary;
1232                         $newvalue =
1233                             preg_replace($valmatch, $valrepl, $attvalue);
1234                         if ($newvalue != $attvalue){
1235                             $attary{$attname} = $newvalue;
1236                         }
1237                     }
1238                 }
1239             }
1240         }
1241         /**
1242          * Turn cid: urls into http-friendly ones.
1243          */
1244         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1245             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1246         }
1247     }
1248     /**
1249      * See if we need to append any attributes to this tag.
1250      */
1251     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1252         if (preg_match($matchtag, $tagname)){
1253             $attary = array_merge($attary, $addattary);
1254         }
1255     }
1256     return $attary;
1257 }
1258
1259 /**
1260  * This function edits the style definition to make them friendly and
1261  * usable in squirrelmail.
1262  *
1263  * @param  $message  the message object
1264  * @param  $id       the message id
1265  * @param  $content  a string with whatever is between <style> and </style>
1266  * @return           a string with edited content.
1267  */
1268 function sq_fixstyle($message, $id, $content){
1269     global $view_unsafe_images;
1270     $me = 'sq_fixstyle';
1271     /**
1272      * First look for general BODY style declaration, which would be
1273      * like so:
1274      * body {background: blah-blah}
1275      * and change it to .bodyclass so we can just assign it to a <div>
1276      */
1277     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1278     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1279     /**
1280      * Fix url('blah') declarations.
1281      */
1282     $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1283                             "url(\\1$secremoveimg\\2)", $content);
1284     /**
1285      * Fix url('https*://.*) declarations but only if $view_unsafe_images
1286      * is false.
1287      */
1288     if (!$view_unsafe_images){
1289         $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1290                                 "url(\\1$secremoveimg\\2)", $content);
1291     }
1292
1293     /**
1294      * Fix urls that refer to cid:
1295      */
1296     while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1297                       $matches)){
1298         $cidurl = $matches{1};
1299         $httpurl = sq_cid2http($message, $id, $cidurl);
1300         $content = preg_replace("|url\($cidurl\)|si",
1301                                 "url($httpurl)", $content);
1302     }
1303
1304     /**
1305      * Fix stupid css declarations which lead to vulnerabilities
1306      * in IE.
1307      */
1308     $match   = Array('/expression/si',
1309                      '/behaviou*r/si',
1310                      '/binding/si');
1311     $replace = Array('idiocy', 'idiocy', 'idiocy');
1312     $content = preg_replace($match, $replace, $content);
1313     return $content;
1314 }
1315
1316 /**
1317  * This function converts cid: url's into the ones that can be viewed in
1318  * the browser.
1319  *
1320  * @param  $message  the message object
1321  * @param  $id       the message id
1322  * @param  $cidurl   the cid: url.
1323  * @return           a string with a http-friendly url
1324  */
1325 function sq_cid2http($message, $id, $cidurl, $mailbox){
1326     /**
1327      * Get rid of quotes.
1328      */
1329     $quotchar = substr($cidurl, 0, 1);
1330     $cidurl = str_replace($quotchar, "", $cidurl);
1331     $cidurl = substr(trim($cidurl), 4);
1332     $linkurl = find_ent_id($cidurl, $message);
1333     /* in case of non-save cid links $httpurl should be replaced by a sort of
1334        unsave link image */
1335     $httpurl = '';
1336     if ($linkurl) {
1337         $httpurl = $quotchar . '../src/download.php?absolute_dl=true&amp;' .
1338                    "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1339                    '&amp;ent_id=' . $linkurl . $quotchar;
1340     }
1341     return $httpurl;
1342 }
1343
1344 /**
1345  * This function changes the <body> tag into a <div> tag since we
1346  * can't really have a body-within-body.
1347  *
1348  * @param  $attary  an array of attributes and values of <body>
1349  * @return          a modified array of attributes to be set for <div>
1350  */
1351 function sq_body2div($attary){
1352     $me = 'sq_body2div';
1353     $divattary = Array('class' => "'bodyclass'");
1354     $bgcolor = '#ffffff';
1355     $text = '#000000';
1356     $styledef = '';
1357     if (is_array($attary) && sizeof($attary) > 0){
1358         foreach ($attary as $attname=>$attvalue){
1359             $quotchar = substr($attvalue, 0, 1);
1360             $attvalue = str_replace($quotchar, "", $attvalue);
1361             switch ($attname){
1362                 case 'background':
1363                     $styledef .= "background-image: url('$attvalue'); ";
1364                     break;
1365                 case 'bgcolor':
1366                     $styledef .= "background-color: $attvalue; ";
1367                     break;
1368                 case 'text':
1369                     $styledef .= "color: $attvalue; ";
1370                     break;
1371             }
1372         }
1373         if (strlen($styledef) > 0){
1374             $divattary{"style"} = "\"$styledef\"";
1375         }
1376     }
1377     return $divattary;
1378 }
1379
1380 /**
1381  * This is the main function and the one you should actually be calling.
1382  * There are several variables you should be aware of an which need
1383  * special description.
1384  *
1385  * Since the description is quite lengthy, see it here:
1386  * http://www.mricon.com/html/phpfilter.html
1387  *
1388  * @param $body                 the string with HTML you wish to filter
1389  * @param $tag_list             see description above
1390  * @param $rm_tags_with_content see description above
1391  * @param $self_closing_tags    see description above
1392  * @param $force_tag_closing    see description above
1393  * @param $rm_attnames          see description above
1394  * @param $bad_attvals          see description above
1395  * @param $add_attr_to_tag      see description above
1396  * @param $message              message object
1397  * @param $id                   message id
1398  * @return                      sanitized html safe to show on your pages.
1399  */
1400 function sq_sanitize($body,
1401                      $tag_list,
1402                      $rm_tags_with_content,
1403                      $self_closing_tags,
1404                      $force_tag_closing,
1405                      $rm_attnames,
1406                      $bad_attvals,
1407                      $add_attr_to_tag,
1408                      $message,
1409                      $id,
1410                      $mailbox
1411                      ){
1412     $me = 'sq_sanitize';
1413     /**
1414      * Normalize rm_tags and rm_tags_with_content.
1415      */
1416     @array_walk($rm_tags, 'sq_casenormalize');
1417     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1418     @array_walk($self_closing_tags, 'sq_casenormalize');
1419     /**
1420      * See if tag_list is of tags to remove or tags to allow.
1421      * false  means remove these tags
1422      * true   means allow these tags
1423      */
1424     $rm_tags = array_shift($tag_list);
1425     $curpos = 0;
1426     $open_tags = Array();
1427     $trusted = "<!-- begin sanitized html -->\n";
1428     $skip_content = false;
1429     /**
1430      * Take care of netscape's stupid javascript entities like
1431      * &{alert('boo')};
1432      */
1433     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1434
1435     while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1436         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1437         $free_content = substr($body, $curpos, $lt-$curpos);
1438         /**
1439          * Take care of <style>
1440          */
1441         if ($tagname == "style" && $tagtype == 2){
1442             /**
1443              * This is a closing </style>. Edit the
1444              * content before we apply it.
1445              */
1446             $free_content = sq_fixstyle($message, $id, $free_content);
1447         }
1448         if ($skip_content == false){
1449             $trusted .= $free_content;
1450         }
1451         if ($tagname != FALSE){
1452             if ($tagtype == 2){
1453                 if ($skip_content == $tagname){
1454                     /**
1455                      * Got to the end of tag we needed to remove.
1456                      */
1457                     $tagname = false;
1458                     $skip_content = false;
1459                 } else {
1460                     if ($skip_content == false){
1461                         if ($tagname == "body"){
1462                             $tagname = "div";
1463                         } else {
1464                             if (isset($open_tags{$tagname}) &&
1465                                 $open_tags{$tagname} > 0){
1466                                 $open_tags{$tagname}--;
1467                             } else {
1468                                 $tagname = false;
1469                             }
1470                         }
1471                     }
1472                 }
1473             } else {
1474                 /**
1475                  * $rm_tags_with_content
1476                  */
1477                 if ($skip_content == false){
1478                     /**
1479                      * See if this is a self-closing type and change
1480                      * tagtype appropriately.
1481                      */
1482                     if ($tagtype == 1
1483                         && in_array($tagname, $self_closing_tags)){
1484                         $tagtype=3;
1485                     }
1486                     /**
1487                      * See if we should skip this tag and any content
1488                      * inside it.
1489                      */
1490                     if ($tagtype == 1 &&
1491                         in_array($tagname, $rm_tags_with_content)){
1492                         $skip_content = $tagname;
1493                     } else {
1494                         if (($rm_tags == false
1495                              && in_array($tagname, $tag_list)) ||
1496                             ($rm_tags == true &&
1497                              !in_array($tagname, $tag_list))){
1498                             $tagname = false;
1499                         } else {
1500                             if ($tagtype == 1){
1501                                 if (isset($open_tags{$tagname})){
1502                                     $open_tags{$tagname}++;
1503                                 } else {
1504                                     $open_tags{$tagname}=1;
1505                                 }
1506                             }
1507                             /**
1508                              * This is where we run other checks.
1509                              */
1510                             if (is_array($attary) && sizeof($attary) > 0){
1511                                 $attary = sq_fixatts($tagname,
1512                                                      $attary,
1513                                                      $rm_attnames,
1514                                                      $bad_attvals,
1515                                                      $add_attr_to_tag,
1516                                                      $message,
1517                                                      $id,
1518                                                      $mailbox
1519                                                      );
1520                             }
1521                             /**
1522                              * Convert body into div.
1523                              */
1524                             if ($tagname == "body"){
1525                                 $tagname = "div";
1526                                 $attary = sq_body2div($attary, $message, $id);
1527                             }
1528                         }
1529                     }
1530                 }
1531             }
1532             if ($tagname != false && $skip_content == false){
1533                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1534             }
1535         }
1536         $curpos = $gt+1;
1537     }
1538     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1539     if ($force_tag_closing == true){
1540         foreach ($open_tags as $tagname=>$opentimes){
1541             while ($opentimes > 0){
1542                 $trusted .= '</' . $tagname . '>';
1543                 $opentimes--;
1544             }
1545         }
1546         $trusted .= "\n";
1547     }
1548     $trusted .= "<!-- end sanitized html -->\n";
1549     return $trusted;
1550 }
1551
1552 /**
1553  * This is a wrapper function to call html sanitizing routines.
1554  *
1555  * @param  $body  the body of the message
1556  * @param  $id    the id of the message
1557  * @return        a string with html safe to display in the browser.
1558  */
1559 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1560     global $attachment_common_show_images, $view_unsafe_images,
1561            $has_unsafe_images;
1562     /**
1563      * Don't display attached images in HTML mode.
1564      */
1565     $attachment_common_show_images = false;
1566     $tag_list = Array(
1567                       false,
1568                       "object",
1569                       "meta",
1570                       "html",
1571                       "head",
1572                       "base",
1573                       "link",
1574                       "frame",
1575                       "iframe"
1576                       );
1577
1578     $rm_tags_with_content = Array(
1579                                   "script",
1580                                   "applet",
1581                                   "embed",
1582                                   "title"
1583                                   );
1584
1585     $self_closing_tags =  Array(
1586                                 "img",
1587                                 "br",
1588                                 "hr",
1589                                 "input"
1590                                 );
1591
1592     $force_tag_closing = false;
1593
1594     $rm_attnames = Array(
1595                          "/.*/" =>
1596                          Array(
1597                                "/target/si",
1598                                "/^on.*/si",
1599                                "/^dynsrc/si",
1600                                "/^data.*/si"
1601                                )
1602                          );
1603
1604     $secremoveimg = "../images/" . _("sec_remove_eng.png");
1605     $bad_attvals = Array(
1606         "/.*/" =>
1607             Array(
1608                 "/^src|background/i" =>
1609                     Array(
1610                           Array(
1611                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1612                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1613                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1614                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1615                                 ),
1616                           Array(
1617                                 "\\1$secremoveimg\\2",
1618                                 "\\1$secremoveimg\\2",
1619                                 "\\1$secremoveimg\\2",
1620                                 "\\1$secremoveimg\\2"
1621                                 )
1622                         ),
1623                 "/^href|action/i" =>
1624                     Array(
1625                           Array(
1626                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1627                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1628                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1629                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1630                                 ),
1631                           Array(
1632                                 "\\1#\\2",
1633                                 "\\1#\\2",
1634                                 "\\1#\\2",
1635                                 "\\1#\\2"
1636                                 )
1637                         ),
1638                 "/^style/si" =>
1639                     Array(
1640                           Array(
1641                                 "/expression/si",
1642                                 "/binding/si",
1643                                 "/behaviou*r/si",
1644                                 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1645                                 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1646                                 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1647                                 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1648                                ),
1649                           Array(
1650                                 "idiocy",
1651                                 "idiocy",
1652                                 "idiocy",
1653                                 "url(\\1#\\2)",
1654                                 "url(\\1#\\2)",
1655                                 "url(\\1#\\2)",
1656                                 "url(\\1#\\2)"
1657                                )
1658                           )
1659                 )
1660         );
1661     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1662         $view_unsafe_images = false;
1663     }
1664     if (!$view_unsafe_images){
1665         /**
1666          * Remove any references to http/https if view_unsafe_images set
1667          * to false.
1668          */
1669          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1670                     '/^([\'\"])\s*https*:.*([\'\"])/si');
1671          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1672                     "\\1$secremoveimg\\2");
1673          array_push($bad_attvals{'/.*/'}{'/^style/si'}[0],
1674                     '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1675          array_push($bad_attvals{'/.*/'}{'/^style/si'}[1],
1676                     "url(\\1$secremoveimg\\2)");
1677     }
1678
1679     $add_attr_to_tag = Array(
1680                              "/^a$/si" => Array('target'=>'"_new"')
1681                              );
1682     $trusted = sq_sanitize($body,
1683                            $tag_list,
1684                            $rm_tags_with_content,
1685                            $self_closing_tags,
1686                            $force_tag_closing,
1687                            $rm_attnames,
1688                            $bad_attvals,
1689                            $add_attr_to_tag,
1690                            $message,
1691                            $id,
1692                            $mailbox
1693                            );
1694     if (preg_match("|$secremoveimg|si", $trusted)){
1695         $has_unsafe_images = true;
1696     }
1697     return $trusted;
1698 }
1699
1700 ?>