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         /* removed urldecode because $_GET is auto urldecoded ??? */
  38         displayPageHeader( $color, $mailbox );
  39         echo "<BODY TEXT=\"$color[8]\" BGCOLOR=\"$color[4]\" LINK=\"$color[7]\" VLINK=\"$color[7]\" ALINK=\"$color[7]\">\n\n" .
  40          '<CENTER>';
  41         $errormessage  = _("SquirrelMail could not decode the bodystructure of the message");
  42         $errormessage .= '<BR>'._("the provided bodystructure by your imap-server").':<BR><BR>';
  43         $errormessage .= '<table><tr><td>' . htmlspecialchars($read) . '</td></tr></table>';
  44         plain_error_message( $errormessage, $color );
  45         echo '</body></html>';
  46         exit;
  47     }
  48     if (count($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=1) {
  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         $cmd = "FETCH $id BODY[]";
 104     } else {
 105         $cmd = "FETCH $id BODY[$ent_id]";
 106     }
 107
 108     $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, $uid_support);
 109     do {
 110         $topline = trim(array_shift($data));
 111     } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
 112
 113     $wholemessage = implode('', $data);
 114     if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
 115         $ret = substr($wholemessage, 0, $regs[1]);
 116         /* There is some information in the content info header that could be important
 117          * in order to parse html messages. Let's get them here.
 118          */
 119         if ($ret{0} == '<') {
 120             $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, $uid_support);
 121         }
 122     } else if (ereg('"([^"]*)"', $topline, $regs)) {
 123         $ret = $regs[1];
 124     } else {
 125         global $where, $what, $mailbox, $passed_id, $startMessage;
 126         $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
 127         if (isset($where) && isset($what)) {
 128             $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
 129         } else {
 130             $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 131         }
 132         $par .= '&amp;response=' . urlencode($response) .
 133                 '&amp;message='  . urlencode($message)  .
 134                 '&amp;topline='  . urlencode($topline);
 135
 136         echo   '<tt><br>' .
 137                '<table width="80%"><tr>' .
 138                '<tr><td colspan=2>' .
 139                _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
 140                '</td></tr>' .
 141                '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
 142                '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
 143                '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
 144                '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
 145                "</table><BR></tt></font><hr>";
 146
 147         $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, $uid_support);
 148         array_shift($data);
 149         $wholemessage = implode('', $data);
 150
 151         $ret = $wholemessage;
 152     }
 153     return $ret;
 154 }
 155
 156 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding) {
 157     global $uid_support;
 158
 159     $sid = sqimap_session_id($uid_support);
 160     /* Don't kill the connection if the browser is over a dialup
 161      * and it would take over 30 seconds to download it.
 162      * Don´t call set_time_limit in safe mode.
 163      */
 164
 165     if (!ini_get('safe_mode')) {
 166         set_time_limit(0);
 167     }
 168     if ($uid_support) {
 169        $sid_s = substr($sid,0,strpos($sid, ' '));
 170     } else {
 171        $sid_s = $sid;
 172     }
 173
 174     $body = mime_fetch_body ($imap_stream, $id, $ent_id);
 175     echo decodeBody($body, $encoding);
 176     return;
 177 /*
 178     fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
 179     $cnt = 0;
 180     $continue = true;
 181     $read = fgets ($imap_stream,8192);
 182
 183
 184     // This could be bad -- if the section has sqimap_session_id() . ' OK'
 185     // or similar, it will kill the download.
 186     while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
 187         if (trim($read) == ')==') {
 188             $read1 = $read;
 189             $read = fgets ($imap_stream,4096);
 190             if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
 191                 return;
 192             } else {
 193                 echo decodeBody($read1, $encoding) .
 194                      decodeBody($read, $encoding);
 195             }
 196         } else if ($cnt) {
 197             echo decodeBody($read, $encoding);
 198         }
 199         $read = fgets ($imap_stream,4096);
 200         $cnt++;
 201 //      break;
 202     }
 203 */
 204 }
 205
 206 /* -[ END MIME DECODING ]----------------------------------------------------------- */
 207
 208 /* This is here for debugging purposes.  It will print out a list
 209  * of all the entity IDs that are in the $message object.
 210  */
 211 function listEntities ($message) {
 212     if ($message) {
 213         echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br>';
 214         for ($i = 0; isset($message->entities[$i]); $i++) {
 215             echo "$i : ";
 216             $msg = listEntities($message->entities[$i]);
 217
 218             if ($msg) {
 219                 echo "return: ";
 220                 return $msg;
 221             }
 222         }
 223     }
 224 }
 225
 226 function getPriorityStr($priority) {
 227     $priority_level = substr($priority,0,1);
 228
 229     switch($priority_level) {
 230         /* Check for a higher then normal priority. */
 231         case '1':
 232         case '2':
 233             $priority_string = _("High");
 234             break;
 235
 236         /* Check for a lower then normal priority. */
 237         case '4':
 238         case '5':
 239             $priority_string = _("Low");
 240             break;
 241
 242         /* Check for a normal priority. */
 243         case '3':
 244         default:
 245             $priority_level = '3';
 246             $priority_string = _("Normal");
 247             break;
 248
 249     }
 250     return $priority_string;
 251 }
 252
 253 /* returns a $message object for a particular entity id */
 254 function getEntity ($message, $ent_id) {
 255     return $message->getEntity($ent_id);
 256 }
 257
 258 /* translateText
 259  * Extracted from strings.php 23/03/2002
 260  */
 261
 262 function translateText(&$body, $wrap_at, $charset) {
 263     global $where, $what;   /* from searching */
 264     global $color;          /* color theme */
 265
 266     require_once(SM_PATH . 'functions/url_parser.php');
 267
 268     $body_ary = explode("\n", $body);
 269     for ($i=0; $i < count($body_ary); $i++) {
 270         $line = $body_ary[$i];
 271         if (strlen($line) - 2 >= $wrap_at) {
 272             sqWordWrap($line, $wrap_at);
 273         }
 274         $line = charset_decode($charset, $line);
 275         $line = str_replace("\t", '        ', $line);
 276
 277         parseUrl ($line);
 278
 279         $quotes = 0;
 280         $pos = 0;
 281         $j = strlen($line);
 282
 283         while ($pos < $j) {
 284             if ($line[$pos] == ' ') {
 285                 $pos++;
 286             } else if (strpos($line, '&gt;', $pos) === $pos) {
 287                 $pos += 4;
 288                 $quotes++;
 289             } else {
 290                 break;
 291             }
 292         }
 293
 294         if ($quotes % 2) {
 295             if (!isset($color[13])) {
 296                 $color[13] = '#800000';
 297             }
 298             $line = '<font color="' . $color[13] . '">' . $line . '</font>';
 299         } elseif ($quotes) {
 300             if (!isset($color[14])) {
 301                 $color[14] = '#FF0000';
 302             }
 303             $line = '<font color="' . $color[14] . '">' . $line . '</font>';
 304         }
 305
 306         $body_ary[$i] = $line;
 307     }
 308     $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
 309 }
 310
 311 /* This returns a parsed string called $body. That string can then
 312  * be displayed as the actual message in the HTML. It contains
 313  * everything needed, including HTML Tags, Attachments at the
 314  * bottom, etc.
 315  */
 316 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
 317     /* This if statement checks for the entity to show as the
 318      * primary message. To add more of them, just put them in the
 319      * order that is their priority.
 320      */
 321     global $startMessage, $username, $key, $imapServerAddress, $imapPort,
 322            $show_html_default, $sort, $has_unsafe_images, $passed_ent_id;
 323
 324     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
 325         $view_unsafe_images = false;
 326     }
 327
 328     $body = '';
 329     $urlmailbox = urlencode($mailbox);
 330     $body_message = getEntity($message, $ent_num);
 331     if (($body_message->header->type0 == 'text') ||
 332         ($body_message->header->type0 == 'rfc822')) {
 333         $body = mime_fetch_body ($imap_stream, $id, $ent_num);
 334         $body = decodeBody($body, $body_message->header->encoding);
 335         $hookResults = do_hook("message_body", $body);
 336         $body = $hookResults[1];
 337
 338         /* If there are other types that shouldn't be formatted, add
 339          * them here.
 340          */
 341
 342         if ($body_message->header->type1 == 'html') {
 343             if ($show_html_default <> 1) {
 344                 $entity_conv = array('&nbsp;' => ' ',
 345                                      '<p>'    => "\n",
 346                                      '<br>'   => "\n",
 347                                      '<P>'    => "\n",
 348                                      '<BR>'   => "\n",
 349                                      '&gt;'   => '>',
 350                                      '&lt;'   => '<');
 351                 $body = strtr($body, $entity_conv);
 352                 $body = strip_tags($body);
 353                 $body = trim($body);
 354                 translateText($body, $wrap_at,
 355                               $body_message->header->getParameter('charset'));
 356             } else {
 357                 $body = magicHTML($body, $id, $message, $mailbox);
 358             }
 359         } else {
 360             translateText($body, $wrap_at,
 361                           $body_message->header->getParameter('charset'));
 362         }
 363         $link = 'read_body.php?passed_id=' . $id . '&amp;ent_id='.$ent_num.
 364                  '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
 365                  '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 366         if (isset($passed_ent_id)) {
 367             $link .= '&amp;passed_ent_id='.$passed_ent_id;
 368         }
 369         if ($view_unsafe_images) {
 370             $text = _("Hide Unsafe Images");
 371         } else {
 372             if (isset($has_unsafe_images) && $has_unsafe_images) {
 373                 $link .= '&amp;view_unsafe_images=1';
 374                 $text = _("View Unsafe Images");
 375             } else {
 376                 $text = '';
 377             }
 378         }
 379         $body .= '<center><small><a href="'.$link.'">'.$text.
 380                  '</a></small></center><br>' . "\n";
 381     }
 382     return $body;
 383 }
 384
 385
 386 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 387     global $where, $what, $startMessage, $color;
 388     static $ShownHTML = 0;
 389
 390     $att_ar = $message->getAttachments($exclude_id);
 391
 392     if (!count($att_ar)) return '';
 393
 394     $attachments = '';
 395
 396     $urlMailbox = urlencode($mailbox);
 397
 398     foreach ($att_ar as $att) {
 399         $ent = $att->entity_id;
 400         $header = $att->header;
 401         $type0 = strtolower($header->type0);
 402         $type1 = strtolower($header->type1);
 403         $name = '';
 404         $links['download link']['text'] = _("download");
 405         $links['download link']['href'] =
 406                 "../src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 407         $ImageURL = '';
 408         if ($type0 =='message' && $type1 == 'rfc822') {
 409             $default_page = '../src/read_body.php';
 410             $rfc822_header = $att->rfc822_header;
 411             $filename = $rfc822_header->subject;
 412             if (trim( $filename ) == '') {
 413                 $filename = 'untitled-[' . $ent . ']' ;
 414             }
 415             $from_o = $rfc822_header->from;
 416             if (is_object($from_o)) {
 417                 $from_name = $from_o->getAddress(false);
 418             } else {
 419                 $from_name = _("Unknown sender");
 420             }
 421             $from_name = decodeHeader(($from_name));
 422             $description = $from_name;
 423         } else {
 424             $default_page = '../src/download.php';
 425             if (is_object($header->disposition)) {
 426                 $filename = $header->disposition->getProperty('filename');
 427                 if (trim($filename) == '') {
 428                     $name = decodeHeader($header->disposition->getProperty('name'));
 429                     if (trim($name) == '') {
 430                         $name = $header->getParameter('name');
 431                         if(trim($name) == '') {
 432                             if (trim( $header->id ) == '') {
 433                                 $filename = 'untitled-[' . $ent . ']' ;
 434                             } else {
 435                                 $filename = 'cid: ' . $header->id;
 436                             }
 437                         } else {
 438                             $filename = $name;
 439                         }
 440                     } else {
 441                         $filename = $name;
 442                     }
 443                 }
 444             } else {
 445                 $filename = $header->getParameter('name');
 446                 if (!trim($filename)) {
 447                     if (trim( $header->id ) == '') {
 448                         $filename = 'untitled-[' . $ent . ']' ;
 449                     } else {
 450                         $filename = 'cid: ' . $header->id;
 451                     }
 452                 }
 453             }
 454             if ($header->description) {
 455                 $description = decodeHeader($header->description);
 456             } else {
 457                 $description = '';
 458             }
 459         }
 460
 461         $display_filename = $filename;
 462         if (isset($passed_ent_id)) {
 463             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 464         } else {
 465             $passed_ent_id_link = '';
 466         }
 467         $defaultlink = $default_page . "?startMessage=$startMessage"
 468                      . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 469                      . '&amp;ent_id='.$ent.$passed_ent_id_link.'&amp;absolute_dl=true';
 470         if ($where && $what) {
 471            $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 472         }
 473         /* This executes the attachment hook with a specific MIME-type.
 474          * If that doesn't have results, it tries if there's a rule
 475          * for a more generic type.
 476          */
 477         $hookresults = do_hook("attachment $type0/$type1", $links,
 478                                $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 479                                $display_filename, $where, $what);
 480         if(count($hookresults[1]) <= 1) {
 481             $hookresults = do_hook("attachment $type0/*", $links,
 482                                    $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 483                                    $display_filename, $where, $what);
 484         }
 485
 486         $links = $hookresults[1];
 487         $defaultlink = $hookresults[6];
 488
 489         $attachments .= '<TR><TD>' .
 490                         '<A HREF="'.$defaultlink.'">'.decodeHeader($display_filename).'</A>&nbsp;</TD>' .
 491                         '<TD><SMALL><b>' . show_readable_size($header->size) .
 492                         '</b>&nbsp;&nbsp;</small></TD>' .
 493                         "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
 494                         '<TD><SMALL>';
 495         $attachments .= '<b>' . $description . '</b>';
 496         $attachments .= '</SMALL></TD><TD><SMALL>&nbsp;';
 497
 498         $skipspaces = 1;
 499         foreach ($links as $val) {
 500             if ($skipspaces) {
 501                 $skipspaces = 0;
 502             } else {
 503                 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
 504             }
 505             $attachments .= '<a href="' . $val['href'] . '">' .  $val['text'] . '</a>';
 506         }
 507         unset($links);
 508         $attachments .= "</TD></TR>\n";
 509     }
 510     return $attachments;
 511 }
 512
 513 /* This function decodes the body depending on the encoding type. */
 514 function decodeBody($body, $encoding) {
 515     global $languages, $squirrelmail_language;
 516     global $show_html_default;
 517
 518     $body = str_replace("\r\n", "\n", $body);
 519     $encoding = strtolower($encoding);
 520
 521     if ($encoding == 'quoted-printable' ||
 522         $encoding == 'quoted_printable') {
 523         $body = quoted_printable_decode($body);
 524
 525         while (ereg("=\n", $body)) {
 526             $body = ereg_replace ("=\n", '', $body);
 527         }
 528
 529     } else if ($encoding == 'base64') {
 530         $body = base64_decode($body);
 531     }
 532
 533     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 534         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 535         $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
 536     }
 537
 538     // All other encodings are returned raw.
 539     return $body;
 540 }
 541
 542 /*
 543  * This functions decode strings that is encoded according to
 544  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 545  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 546  */
 547 function decodeHeader ($string, $utfencode=true,$htmlsave=true) {
 548     global $languages, $squirrelmail_language;
 549     if (is_array($string)) {
 550         $string = implode("\n", $string);
 551     }
 552
 553     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 554         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 555         $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
 556         // Do we need to return at this point?
 557         // return $string;
 558     }
 559     $i = 0;
 560     $aString = explode(' ',$string);
 561     foreach ($aString as $chunk) {
 562         $encoded = false;
 563         $aString[$i] = '';
 564         while (preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 565             //$aString[$i] .= $res[1];
 566             //echo "$chunk match ". $res[5] . "<br>";
 567             $encoding = ucfirst($res[3]);
 568             switch ($encoding)
 569             {
 570             case 'B':
 571                 $replace = base64_decode($res[4]);
 572                 $aString[$i] .= charset_decode($res[2],$replace);
 573                 break;
 574             case 'Q':
 575                 $replace = str_replace('_', ' ', $res[4]);
 576                 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 577                                     $replace);
 578                 /* Only encode into entities by default. Some places
 579                  * don't need the encoding, like the compose form.
 580                  */
 581                 if ($utfencode) {
 582                     $replace = charset_decode($res[2], $replace);
 583                 } else {
 584                     if ($htmlsave) {
 585                         $replace = htmlspecialchars($replace);
 586                     }
 587                 }
 588                 $aString[$i] .= $replace;
 589                 break;
 590             default:
 591                 break;
 592             }
 593             $chunk = $res[5];
 594             $encoded = true;
 595         }
 596         if (!$encoded && $htmlsave) {
 597             $aString[$i] = htmlspecialchars($chunk);
 598         } else {
 599             $aString[$i] .= $chunk;
 600         }
 601         ++$i;
 602     }
 603     return implode (' ',$aString);
 604
 605 }
 606
 607 /*
 608  * Encode a string according to RFC 1522 for use in headers if it
 609  * contains 8-bit characters or anything that looks like it should
 610  * be encoded.
 611  */
 612 function encodeHeader ($string) {
 613     global $default_charset, $languages, $squirrelmail_language;
 614
 615     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 616         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 617         return  $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
 618     }
 619     if (strtolower($default_charset) == 'iso-8859-1') {
 620         $string = str_replace("\240",' ',$string);
 621     }
 622
 623     // Encode only if the string contains 8-bit characters or =?
 624     $j = strlen($string);
 625     $max_l = 75 - strlen($default_charset) - 7;
 626     $aRet = array();
 627     $ret = '';
 628     $iEncStart = $enc_init = false;
 629     $cur_l = $iOffset = 0;
 630     for($i = 0; $i < $j; ++$i) {
 631         switch($string{$i})
 632         {
 633         case '=':
 634         case '<':
 635         case '>':
 636         case ',':
 637         case '?':
 638         case '_':
 639             if ($iEncStart === false) {
 640                 $iEncStart = $i;
 641             }
 642             $cur_l+=3;
 643             if ($cur_l > ($max_l-2)) {
 644                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 645                 $aRet[] = "=?$default_charset?Q?$ret?=";
 646                 $iOffset = $i;
 647                 $cur_l = 0;
 648                 $ret = '';
 649                 $iEncStart = false;
 650             } else {
 651                 $ret .= sprintf("=%02X",ord($string{$i}));
 652             }
 653             break;
 654         case '(':
 655         case ')':
 656             if ($iEncStart !== false) {
 657                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 658                 $aRet[] = "=?$default_charset?Q?$ret?=";
 659                 $iOffset = $i;
 660                 $cur_l = 0;
 661                 $ret = '';
 662                 $iEncStart = false;
 663             }
 664             break;
 665         case ' ':
 666             if ($iEncStart !== false) {
 667                 $cur_l++;
 668                 if ($cur_l > $max_l) {
 669                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 670                     $aRet[] = "=?$default_charset?Q?$ret?=";
 671                     $iOffset = $i;
 672                     $cur_l = 0;
 673                     $ret = '';
 674                     $iEncStart = false;
 675                 } else {
 676                     $ret .= '_';
 677                 }
 678             }
 679             break;
 680         default:
 681             $k = ord($string{$i});
 682             if ($k > 126) {
 683                 if ($iEncStart === false) {
 684                     $iEncStart = $i;
 685                 }
 686                 $cur_l += 3;
 687                 if ($cur_l > ($max_l-2)) {
 688                     if ($iEncStart !== false) {
 689                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 690                         $aRet[] = "=?$default_charset?Q?$ret?=";
 691                     } else {
 692                         $aRet[] = substr($string,$iOffset,$i-$iOffset);
 693                     }
 694                     $cur_l = 3;
 695                     $ret = '';
 696                     $iOffset = $i;
 697                 }
 698                 $enc_init = true;
 699                 $ret .= sprintf("=%02X", $k);
 700             } else {
 701                 if ($iEncStart !== false) {
 702                     $cur_l++;
 703                     if ($cur_l > $max_l) {
 704                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 705                         $aRet[] = "=?$default_charset?Q?$ret?=";
 706                         $iEncStart = false;
 707                         $iOffset = $i;
 708                         $cur_l = 0;
 709                         $ret = '';
 710                     } else {
 711                         $ret .= $string{$i};
 712                     }
 713                 }
 714             }
 715             break;
 716         }
 717     }
 718
 719     if ($enc_init) {
 720         if ($iEncStart !== false) {
 721             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 722             $aRet[] = "=?$default_charset?Q?$ret?=";
 723         } else {
 724             $aRet[] = substr($string,$iOffset);
 725         }
 726         $string = implode('',$aRet);
 727     }
 728     return $string;
 729 }
 730
 731 /* This function trys to locate the entity_id of a specific mime element */
 732 function find_ent_id($id, $message) {
 733     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
 734         if ($message->entities[$i]->header->type0 == 'multipart')  {
 735             $ret = find_ent_id($id, $message->entities[$i]);
 736         } else {
 737             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
 738 //                if (sq_check_save_extension($message->entities[$i])) {
 739                     return $message->entities[$i]->entity_id;
 740 //                }
 741             }
 742         }
 743     }
 744     return $ret;
 745 }
 746
 747 function sq_check_save_extension($message) {
 748     $filename = $message->getFilename();
 749     $ext = substr($filename, strrpos($filename,'.')+1);
 750     $save_extensions = array('jpg','jpeg','gif','png','bmp');
 751     return in_array($ext, $save_extensions);
 752 }
 753
 754
 755 /**
 756  ** HTMLFILTER ROUTINES
 757  */
 758
 759 /**
 760  * This function returns the final tag out of the tag name, an array
 761  * of attributes, and the type of the tag. This function is called by
 762  * sq_sanitize internally.
 763  *
 764  * @param  $tagname  the name of the tag.
 765  * @param  $attary   the array of attributes and their values
 766  * @param  $tagtype  The type of the tag (see in comments).
 767  * @return           a string with the final tag representation.
 768  */
 769 function sq_tagprint($tagname, $attary, $tagtype){
 770     $me = 'sq_tagprint';
 771
 772     if ($tagtype == 2){
 773         $fulltag = '</' . $tagname . '>';
 774     } else {
 775         $fulltag = '<' . $tagname;
 776         if (is_array($attary) && sizeof($attary)){
 777             $atts = Array();
 778             while (list($attname, $attvalue) = each($attary)){
 779                 array_push($atts, "$attname=$attvalue");
 780             }
 781             $fulltag .= ' ' . join(" ", $atts);
 782         }
 783         if ($tagtype == 3){
 784             $fulltag .= ' /';
 785         }
 786         $fulltag .= '>';
 787     }
 788     return $fulltag;
 789 }
 790
 791 /**
 792  * A small helper function to use with array_walk. Modifies a by-ref
 793  * value and makes it lowercase.
 794  *
 795  * @param  $val a value passed by-ref.
 796  * @return      void since it modifies a by-ref value.
 797  */
 798 function sq_casenormalize(&$val){
 799     $val = strtolower($val);
 800 }
 801
 802 /**
 803  * This function skips any whitespace from the current position within
 804  * a string and to the next non-whitespace value.
 805  *
 806  * @param  $body   the string
 807  * @param  $offset the offset within the string where we should start
 808  *                 looking for the next non-whitespace character.
 809  * @return         the location within the $body where the next
 810  *                 non-whitespace char is located.
 811  */
 812 function sq_skipspace($body, $offset){
 813     $me = 'sq_skipspace';
 814     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
 815     if (sizeof($matches{1})){
 816         $count = strlen($matches{1});
 817         $offset += $count;
 818     }
 819     return $offset;
 820 }
 821
 822 /**
 823  * This function looks for the next character within a string.  It's
 824  * really just a glorified "strpos", except it catches if failures
 825  * nicely.
 826  *
 827  * @param  $body   The string to look for needle in.
 828  * @param  $offset Start looking from this position.
 829  * @param  $needle The character/string to look for.
 830  * @return         location of the next occurance of the needle, or
 831  *                 strlen($body) if needle wasn't found.
 832  */
 833 function sq_findnxstr($body, $offset, $needle){
 834     $me  = 'sq_findnxstr';
 835     $pos = strpos($body, $needle, $offset);
 836     if ($pos === FALSE){
 837         $pos = strlen($body);
 838     }
 839     return $pos;
 840 }
 841
 842 /**
 843  * This function takes a PCRE-style regexp and tries to match it
 844  * within the string.
 845  *
 846  * @param  $body   The string to look for needle in.
 847  * @param  $offset Start looking from here.
 848  * @param  $reg    A PCRE-style regex to match.
 849  * @return         Returns a false if no matches found, or an array
 850  *                 with the following members:
 851  *                 - integer with the location of the match within $body
 852  *                 - string with whatever content between offset and the match
 853  *                 - string with whatever it is we matched
 854  */
 855 function sq_findnxreg($body, $offset, $reg){
 856     $me = 'sq_findnxreg';
 857     $matches = Array();
 858     $retarr = Array();
 859     preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
 860     if (!$matches{0}){
 861         $retarr = false;
 862     } else {
 863         $retarr{0} = $offset + strlen($matches{1});
 864         $retarr{1} = $matches{1};
 865         $retarr{2} = $matches{2};
 866     }
 867     return $retarr;
 868 }
 869
 870 /**
 871  * This function looks for the next tag.
 872  *
 873  * @param  $body   String where to look for the next tag.
 874  * @param  $offset Start looking from here.
 875  * @return         false if no more tags exist in the body, or
 876  *                 an array with the following members:
 877  *                 - string with the name of the tag
 878  *                 - array with attributes and their values
 879  *                 - integer with tag type (1, 2, or 3)
 880  *                 - integer where the tag starts (starting "<")
 881  *                 - integer where the tag ends (ending ">")
 882  *                 first three members will be false, if the tag is invalid.
 883  */
 884 function sq_getnxtag($body, $offset){
 885     $me = 'sq_getnxtag';
 886     if ($offset > strlen($body)){
 887         return false;
 888     }
 889     $lt = sq_findnxstr($body, $offset, "<");
 890     if ($lt == strlen($body)){
 891         return false;
 892     }
 893     /**
 894      * We are here:
 895      * blah blah <tag attribute="value">
 896      * \---------^
 897      */
 898     $pos = sq_skipspace($body, $lt+1);
 899     if ($pos >= strlen($body)){
 900         return Array(false, false, false, $lt, strlen($body));
 901     }
 902     /**
 903      * There are 3 kinds of tags:
 904      * 1. Opening tag, e.g.:
 905      *    <a href="blah">
 906      * 2. Closing tag, e.g.:
 907      *    </a>
 908      * 3. XHTML-style content-less tag, e.g.:
 909      *    <img src="blah"/>
 910      */
 911     $tagtype = false;
 912     switch (substr($body, $pos, 1)){
 913         case '/':
 914             $tagtype = 2;
 915             $pos++;
 916             break;
 917         case '!':
 918             /**
 919              * A comment or an SGML declaration.
 920              */
 921             if (substr($body, $pos+1, 2) == "--"){
 922                 $gt = strpos($body, "-->", $pos);
 923                 if ($gt === false){
 924                     $gt = strlen($body);
 925                 } else {
 926                     $gt += 2;
 927                 }
 928                 return Array(false, false, false, $lt, $gt);
 929             } else {
 930                 $gt = sq_findnxstr($body, $pos, ">");
 931                 return Array(false, false, false, $lt, $gt);
 932             }
 933             break;
 934         default:
 935             /**
 936              * Assume tagtype 1 for now. If it's type 3, we'll switch values
 937              * later.
 938              */
 939             $tagtype = 1;
 940             break;
 941     }
 942
 943     $tag_start = $pos;
 944     $tagname = '';
 945     /**
 946      * Look for next [\W-_], which will indicate the end of the tag name.
 947      */
 948     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
 949     if ($regary == false){
 950         return Array(false, false, false, $lt, strlen($body));
 951     }
 952     list($pos, $tagname, $match) = $regary;
 953     $tagname = strtolower($tagname);
 954
 955     /**
 956      * $match can be either of these:
 957      * '>'  indicating the end of the tag entirely.
 958      * '\s' indicating the end of the tag name.
 959      * '/'  indicating that this is type-3 xhtml tag.
 960      *
 961      * Whatever else we find there indicates an invalid tag.
 962      */
 963     switch ($match){
 964         case '/':
 965             /**
 966              * This is an xhtml-style tag with a closing / at the
 967              * end, like so: <img src="blah"/>. Check if it's followed
 968              * by the closing bracket. If not, then this tag is invalid
 969              */
 970             if (substr($body, $pos, 2) == "/>"){
 971                 $pos++;
 972                 $tagtype = 3;
 973             } else {
 974                 $gt = sq_findnxstr($body, $pos, ">");
 975                 $retary = Array(false, false, false, $lt, $gt);
 976                 return $retary;
 977             }
 978         case '>':
 979             return Array($tagname, false, $tagtype, $lt, $pos);
 980             break;
 981         default:
 982             /**
 983              * Check if it's whitespace
 984              */
 985             if (!preg_match('/\s/', $match)){
 986                 /**
 987                  * This is an invalid tag! Look for the next closing ">".
 988                  */
 989                 $gt = sq_findnxstr($body, $offset, ">");
 990                 return Array(false, false, false, $lt, $gt);
 991             }
 992             break;
 993     }
 994
 995     /**
 996      * At this point we're here:
 997      * <tagname  attribute='blah'>
 998      * \-------^
 999      *
1000      * At this point we loop in order to find all attributes.
1001      */
1002     $attname = '';
1003     $atttype = false;
1004     $attary = Array();
1005
1006     while ($pos <= strlen($body)){
1007         $pos = sq_skipspace($body, $pos);
1008         if ($pos == strlen($body)){
1009             /**
1010              * Non-closed tag.
1011              */
1012             return Array(false, false, false, $lt, $pos);
1013         }
1014         /**
1015          * See if we arrived at a ">" or "/>", which means that we reached
1016          * the end of the tag.
1017          */
1018         $matches = Array();
1019         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1020             /**
1021              * Yep. So we did.
1022              */
1023             $pos += strlen($matches{1});
1024             if ($matches{2} == "/>"){
1025                 $tagtype = 3;
1026                 $pos++;
1027             }
1028             return Array($tagname, $attary, $tagtype, $lt, $pos);
1029         }
1030
1031         /**
1032          * There are several types of attributes, with optional
1033          * [:space:] between members.
1034          * Type 1:
1035          *   attrname[:space:]=[:space:]'CDATA'
1036          * Type 2:
1037          *   attrname[:space:]=[:space:]"CDATA"
1038          * Type 3:
1039          *   attr[:space:]=[:space:]CDATA
1040          * Type 4:
1041          *   attrname
1042          *
1043          * We leave types 1 and 2 the same, type 3 we check for
1044          * '"' and convert to "&quot" if needed, then wrap in
1045          * double quotes. Type 4 we convert into:
1046          * attrname="yes".
1047          */
1048         $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1049         if ($regary == false){
1050             /**
1051              * Looks like body ended before the end of tag.
1052              */
1053             return Array(false, false, false, $lt, strlen($body));
1054         }
1055         list($pos, $attname, $match) = $regary;
1056         $attname = strtolower($attname);
1057         /**
1058          * We arrived at the end of attribute name. Several things possible
1059          * here:
1060          * '>'  means the end of the tag and this is attribute type 4
1061          * '/'  if followed by '>' means the same thing as above
1062          * '\s' means a lot of things -- look what it's followed by.
1063          *      anything else means the attribute is invalid.
1064          */
1065         switch($match){
1066             case '/':
1067                 /**
1068                  * This is an xhtml-style tag with a closing / at the
1069                  * end, like so: <img src="blah"/>. Check if it's followed
1070                  * by the closing bracket. If not, then this tag is invalid
1071                  */
1072                 if (substr($body, $pos, 2) == "/>"){
1073                     $pos++;
1074                     $tagtype = 3;
1075                 } else {
1076                     $gt = sq_findnxstr($body, $pos, ">");
1077                     $retary = Array(false, false, false, $lt, $gt);
1078                     return $retary;
1079                 }
1080             case '>':
1081                 $attary{$attname} = '"yes"';
1082                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1083                 break;
1084             default:
1085                 /**
1086                  * Skip whitespace and see what we arrive at.
1087                  */
1088                 $pos = sq_skipspace($body, $pos);
1089                 $char = substr($body, $pos, 1);
1090                 /**
1091                  * Two things are valid here:
1092                  * '=' means this is attribute type 1 2 or 3.
1093                  * \w means this was attribute type 4.
1094                  * anything else we ignore and re-loop. End of tag and
1095                  * invalid stuff will be caught by our checks at the beginning
1096                  * of the loop.
1097                  */
1098                 if ($char == "="){
1099                     $pos++;
1100                     $pos = sq_skipspace($body, $pos);
1101                     /**
1102                      * Here are 3 possibilities:
1103                      * "'"  attribute type 1
1104                      * '"'  attribute type 2
1105                      * everything else is the content of tag type 3
1106                      */
1107                     $quot = substr($body, $pos, 1);
1108                     if ($quot == "'"){
1109                         $regary = sq_findnxreg($body, $pos+1, "\'");
1110                         if ($regary == false){
1111                             return Array(false, false, false, $lt, strlen($body));
1112                         }
1113                         list($pos, $attval, $match) = $regary;
1114                         $pos++;
1115                         $attary{$attname} = "'" . $attval . "'";
1116                     } else if ($quot == '"'){
1117                         $regary = sq_findnxreg($body, $pos+1, '\"');
1118                         if ($regary == false){
1119                             return Array(false, false, false, $lt, strlen($body));
1120                         }
1121                         list($pos, $attval, $match) = $regary;
1122                         $pos++;
1123                         $attary{$attname} = '"' . $attval . '"';
1124                     } else {
1125                         /**
1126                          * These are hateful. Look for \s, or >.
1127                          */
1128                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1129                         if ($regary == false){
1130                             return Array(false, false, false, $lt, strlen($body));
1131                         }
1132                         list($pos, $attval, $match) = $regary;
1133                         /**
1134                          * If it's ">" it will be caught at the top.
1135                          */
1136                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1137                         $attary{$attname} = '"' . $attval . '"';
1138                     }
1139                 } else if (preg_match("|[\w/>]|", $char)) {
1140                     /**
1141                      * That was attribute type 4.
1142                      */
1143                     $attary{$attname} = '"yes"';
1144                 } else {
1145                     /**
1146                      * An illegal character. Find next '>' and return.
1147                      */
1148                     $gt = sq_findnxstr($body, $pos, ">");
1149                     return Array(false, false, false, $lt, $gt);
1150                 }
1151                 break;
1152         }
1153     }
1154     /**
1155      * The fact that we got here indicates that the tag end was never
1156      * found. Return invalid tag indication so it gets stripped.
1157      */
1158     return Array(false, false, false, $lt, strlen($body));
1159 }
1160
1161 /**
1162  * This function checks attribute values for entity-encoded values
1163  * and returns them translated into 8-bit strings so we can run
1164  * checks on them.
1165  *
1166  * @param  $attvalue A string to run entity check against.
1167  * @return           Translated value.
1168  */
1169 function sq_deent($attvalue){
1170     $me = 'sq_deent';
1171     /**
1172      * See if we have to run the checks first. All entities must start
1173      * with "&".
1174      */
1175     if (strpos($attvalue, "&") === false){
1176         return $attvalue;
1177     }
1178     /**
1179      * Check named entities first.
1180      */
1181     $trans = get_html_translation_table(HTML_ENTITIES);
1182     /**
1183      * Leave &quot; in, as it can mess us up.
1184      */
1185     $trans = array_flip($trans);
1186     unset($trans{"&quot;"});
1187     while (list($ent, $val) = each($trans)){
1188         $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1189     }
1190     /**
1191      * Now translate numbered entities from 1 to 255 if needed.
1192      */
1193     if (strpos($attvalue, "#") !== false){
1194         $omit = Array(34, 39);
1195         for ($asc=1; $asc<256; $asc++){
1196             if (!in_array($asc, $omit)){
1197                 $chr = chr($asc);
1198                 $attvalue = preg_replace("/\&#0*$asc;*(\D)/si", "$chr\\1",
1199                                          $attvalue);
1200                 $attvalue = preg_replace("/\&#x0*".dechex($asc).";*(\W)/si",
1201                                          "$chr\\1", $attvalue);
1202             }
1203         }
1204     }
1205     return $attvalue;
1206 }
1207
1208 /**
1209  * This function runs various checks against the attributes.
1210  *
1211  * @param  $tagname         String with the name of the tag.
1212  * @param  $attary          Array with all tag attributes.
1213  * @param  $rm_attnames     See description for sq_sanitize
1214  * @param  $bad_attvals     See description for sq_sanitize
1215  * @param  $add_attr_to_tag See description for sq_sanitize
1216  * @param  $message         message object
1217  * @param  $id              message id
1218  * @return                  Array with modified attributes.
1219  */
1220 function sq_fixatts($tagname,
1221                     $attary,
1222                     $rm_attnames,
1223                     $bad_attvals,
1224                     $add_attr_to_tag,
1225                     $message,
1226                     $id,
1227                     $mailbox
1228                     ){
1229     $me = 'sq_fixatts';
1230     while (list($attname, $attvalue) = each($attary)){
1231         /**
1232          * See if this attribute should be removed.
1233          */
1234         foreach ($rm_attnames as $matchtag=>$matchattrs){
1235             if (preg_match($matchtag, $tagname)){
1236                 foreach ($matchattrs as $matchattr){
1237                     if (preg_match($matchattr, $attname)){
1238                         unset($attary{$attname});
1239                         continue;
1240                     }
1241                 }
1242             }
1243         }
1244         /**
1245          * Remove any entities.
1246          */
1247         $attvalue = sq_deent($attvalue);
1248
1249         /**
1250          * Now let's run checks on the attvalues.
1251          * I don't expect anyone to comprehend this. If you do,
1252          * get in touch with me so I can drive to where you live and
1253          * shake your hand personally. :)
1254          */
1255         foreach ($bad_attvals as $matchtag=>$matchattrs){
1256             if (preg_match($matchtag, $tagname)){
1257                 foreach ($matchattrs as $matchattr=>$valary){
1258                     if (preg_match($matchattr, $attname)){
1259                         /**
1260                          * There are two arrays in valary.
1261                          * First is matches.
1262                          * Second one is replacements
1263                          */
1264                         list($valmatch, $valrepl) = $valary;
1265                         $newvalue =
1266                             preg_replace($valmatch, $valrepl, $attvalue);
1267                         if ($newvalue != $attvalue){
1268                             $attary{$attname} = $newvalue;
1269                         }
1270                     }
1271                 }
1272             }
1273         }
1274         /**
1275          * Turn cid: urls into http-friendly ones.
1276          */
1277         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1278             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1279         }
1280     }
1281     /**
1282      * See if we need to append any attributes to this tag.
1283      */
1284     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1285         if (preg_match($matchtag, $tagname)){
1286             $attary = array_merge($attary, $addattary);
1287         }
1288     }
1289     return $attary;
1290 }
1291
1292 /**
1293  * This function edits the style definition to make them friendly and
1294  * usable in squirrelmail.
1295  *
1296  * @param  $message  the message object
1297  * @param  $id       the message id
1298  * @param  $content  a string with whatever is between <style> and </style>
1299  * @return           a string with edited content.
1300  */
1301 function sq_fixstyle($message, $id, $content){
1302     global $view_unsafe_images;
1303     $me = 'sq_fixstyle';
1304     /**
1305      * First look for general BODY style declaration, which would be
1306      * like so:
1307      * body {background: blah-blah}
1308      * and change it to .bodyclass so we can just assign it to a <div>
1309      */
1310     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1311     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1312     /**
1313      * Fix url('blah') declarations.
1314      */
1315     $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1316                             "url(\\1$secremoveimg\\2)", $content);
1317     /**
1318      * Fix url('https*://.*) declarations but only if $view_unsafe_images
1319      * is false.
1320      */
1321     if (!$view_unsafe_images){
1322         $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1323                                 "url(\\1$secremoveimg\\2)", $content);
1324     }
1325
1326     /**
1327      * Fix urls that refer to cid:
1328      */
1329     while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1330                       $matches)){
1331         $cidurl = $matches{1};
1332         $httpurl = sq_cid2http($message, $id, $cidurl);
1333         $content = preg_replace("|url\($cidurl\)|si",
1334                                 "url($httpurl)", $content);
1335     }
1336
1337     /**
1338      * Fix stupid css declarations which lead to vulnerabilities
1339      * in IE.
1340      */
1341     $match   = Array('/expression/si',
1342                      '/behaviou*r/si',
1343                      '/binding/si');
1344     $replace = Array('idiocy', 'idiocy', 'idiocy');
1345     $content = preg_replace($match, $replace, $content);
1346     return $content;
1347 }
1348
1349 /**
1350  * This function converts cid: url's into the ones that can be viewed in
1351  * the browser.
1352  *
1353  * @param  $message  the message object
1354  * @param  $id       the message id
1355  * @param  $cidurl   the cid: url.
1356  * @return           a string with a http-friendly url
1357  */
1358 function sq_cid2http($message, $id, $cidurl, $mailbox){
1359     /**
1360      * Get rid of quotes.
1361      */
1362     $quotchar = substr($cidurl, 0, 1);
1363     $cidurl = str_replace($quotchar, "", $cidurl);
1364     $cidurl = substr(trim($cidurl), 4);
1365     $linkurl = find_ent_id($cidurl, $message);
1366     /* in case of non-save cid links $httpurl should be replaced by a sort of
1367        unsave link image */
1368     $httpurl = '';
1369     if ($linkurl) {
1370         $httpurl = $quotchar . '../src/download.php?absolute_dl=true&amp;' .
1371                    "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1372                    '&amp;ent_id=' . $linkurl . $quotchar;
1373     }
1374     return $httpurl;
1375 }
1376
1377 /**
1378  * This function changes the <body> tag into a <div> tag since we
1379  * can't really have a body-within-body.
1380  *
1381  * @param  $attary  an array of attributes and values of <body>
1382  * @return          a modified array of attributes to be set for <div>
1383  */
1384 function sq_body2div($attary){
1385     $me = 'sq_body2div';
1386     $divattary = Array('class' => "'bodyclass'");
1387     $bgcolor = '#ffffff';
1388     $text = '#000000';
1389     $styledef = '';
1390     if (is_array($attary) && sizeof($attary) > 0){
1391         foreach ($attary as $attname=>$attvalue){
1392             $quotchar = substr($attvalue, 0, 1);
1393             $attvalue = str_replace($quotchar, "", $attvalue);
1394             switch ($attname){
1395                 case 'background':
1396                     $styledef .= "background-image: url('$attvalue'); ";
1397                     break;
1398                 case 'bgcolor':
1399                     $styledef .= "background-color: $attvalue; ";
1400                     break;
1401                 case 'text':
1402                     $styledef .= "color: $attvalue; ";
1403                     break;
1404             }
1405         }
1406         if (strlen($styledef) > 0){
1407             $divattary{"style"} = "\"$styledef\"";
1408         }
1409     }
1410     return $divattary;
1411 }
1412
1413 /**
1414  * This is the main function and the one you should actually be calling.
1415  * There are several variables you should be aware of an which need
1416  * special description.
1417  *
1418  * Since the description is quite lengthy, see it here:
1419  * http://www.mricon.com/html/phpfilter.html
1420  *
1421  * @param $body                 the string with HTML you wish to filter
1422  * @param $tag_list             see description above
1423  * @param $rm_tags_with_content see description above
1424  * @param $self_closing_tags    see description above
1425  * @param $force_tag_closing    see description above
1426  * @param $rm_attnames          see description above
1427  * @param $bad_attvals          see description above
1428  * @param $add_attr_to_tag      see description above
1429  * @param $message              message object
1430  * @param $id                   message id
1431  * @return                      sanitized html safe to show on your pages.
1432  */
1433 function sq_sanitize($body,
1434                      $tag_list,
1435                      $rm_tags_with_content,
1436                      $self_closing_tags,
1437                      $force_tag_closing,
1438                      $rm_attnames,
1439                      $bad_attvals,
1440                      $add_attr_to_tag,
1441                      $message,
1442                      $id,
1443                      $mailbox
1444                      ){
1445     $me = 'sq_sanitize';
1446     /**
1447      * Normalize rm_tags and rm_tags_with_content.
1448      */
1449     @array_walk($rm_tags, 'sq_casenormalize');
1450     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1451     @array_walk($self_closing_tags, 'sq_casenormalize');
1452     /**
1453      * See if tag_list is of tags to remove or tags to allow.
1454      * false  means remove these tags
1455      * true   means allow these tags
1456      */
1457     $rm_tags = array_shift($tag_list);
1458     $curpos = 0;
1459     $open_tags = Array();
1460     $trusted = "<!-- begin sanitized html -->\n";
1461     $skip_content = false;
1462     /**
1463      * Take care of netscape's stupid javascript entities like
1464      * &{alert('boo')};
1465      */
1466     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1467
1468     while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1469         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1470         $free_content = substr($body, $curpos, $lt-$curpos);
1471         /**
1472          * Take care of <style>
1473          */
1474         if ($tagname == "style" && $tagtype == 2){
1475             /**
1476              * This is a closing </style>. Edit the
1477              * content before we apply it.
1478              */
1479             $free_content = sq_fixstyle($message, $id, $free_content);
1480         }
1481         if ($skip_content == false){
1482             $trusted .= $free_content;
1483         }
1484         if ($tagname != FALSE){
1485             if ($tagtype == 2){
1486                 if ($skip_content == $tagname){
1487                     /**
1488                      * Got to the end of tag we needed to remove.
1489                      */
1490                     $tagname = false;
1491                     $skip_content = false;
1492                 } else {
1493                     if ($skip_content == false){
1494                         if ($tagname == "body"){
1495                             $tagname = "div";
1496                         } else {
1497                             if (isset($open_tags{$tagname}) &&
1498                                 $open_tags{$tagname} > 0){
1499                                 $open_tags{$tagname}--;
1500                             } else {
1501                                 $tagname = false;
1502                             }
1503                         }
1504                     }
1505                 }
1506             } else {
1507                 /**
1508                  * $rm_tags_with_content
1509                  */
1510                 if ($skip_content == false){
1511                     /**
1512                      * See if this is a self-closing type and change
1513                      * tagtype appropriately.
1514                      */
1515                     if ($tagtype == 1
1516                         && in_array($tagname, $self_closing_tags)){
1517                         $tagtype=3;
1518                     }
1519                     /**
1520                      * See if we should skip this tag and any content
1521                      * inside it.
1522                      */
1523                     if ($tagtype == 1 &&
1524                         in_array($tagname, $rm_tags_with_content)){
1525                         $skip_content = $tagname;
1526                     } else {
1527                         if (($rm_tags == false
1528                              && in_array($tagname, $tag_list)) ||
1529                             ($rm_tags == true &&
1530                              !in_array($tagname, $tag_list))){
1531                             $tagname = false;
1532                         } else {
1533                             if ($tagtype == 1){
1534                                 if (isset($open_tags{$tagname})){
1535                                     $open_tags{$tagname}++;
1536                                 } else {
1537                                     $open_tags{$tagname}=1;
1538                                 }
1539                             }
1540                             /**
1541                              * This is where we run other checks.
1542                              */
1543                             if (is_array($attary) && sizeof($attary) > 0){
1544                                 $attary = sq_fixatts($tagname,
1545                                                      $attary,
1546                                                      $rm_attnames,
1547                                                      $bad_attvals,
1548                                                      $add_attr_to_tag,
1549                                                      $message,
1550                                                      $id,
1551                                                      $mailbox
1552                                                      );
1553                             }
1554                             /**
1555                              * Convert body into div.
1556                              */
1557                             if ($tagname == "body"){
1558                                 $tagname = "div";
1559                                 $attary = sq_body2div($attary, $message, $id);
1560                             }
1561                         }
1562                     }
1563                 }
1564             }
1565             if ($tagname != false && $skip_content == false){
1566                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1567             }
1568         }
1569         $curpos = $gt+1;
1570     }
1571     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1572     if ($force_tag_closing == true){
1573         foreach ($open_tags as $tagname=>$opentimes){
1574             while ($opentimes > 0){
1575                 $trusted .= '</' . $tagname . '>';
1576                 $opentimes--;
1577             }
1578         }
1579         $trusted .= "\n";
1580     }
1581     $trusted .= "<!-- end sanitized html -->\n";
1582     return $trusted;
1583 }
1584
1585 /**
1586  * This is a wrapper function to call html sanitizing routines.
1587  *
1588  * @param  $body  the body of the message
1589  * @param  $id    the id of the message
1590  * @return        a string with html safe to display in the browser.
1591  */
1592 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1593     global $attachment_common_show_images, $view_unsafe_images,
1594            $has_unsafe_images;
1595     /**
1596      * Don't display attached images in HTML mode.
1597      */
1598     $attachment_common_show_images = false;
1599     $tag_list = Array(
1600                       false,
1601                       "object",
1602                       "meta",
1603                       "html",
1604                       "head",
1605                       "base",
1606                       "link",
1607                       "frame",
1608                       "iframe"
1609                       );
1610
1611     $rm_tags_with_content = Array(
1612                                   "script",
1613                                   "applet",
1614                                   "embed",
1615                                   "title"
1616                                   );
1617
1618     $self_closing_tags =  Array(
1619                                 "img",
1620                                 "br",
1621                                 "hr",
1622                                 "input"
1623                                 );
1624
1625     $force_tag_closing = false;
1626
1627     $rm_attnames = Array(
1628                          "/.*/" =>
1629                          Array(
1630                                "/target/si",
1631                                "/^on.*/si",
1632                                "/^dynsrc/si",
1633                                "/^data.*/si"
1634                                )
1635                          );
1636
1637     $secremoveimg = "../images/" . _("sec_remove_eng.png");
1638     $bad_attvals = Array(
1639         "/.*/" =>
1640             Array(
1641                 "/^src|background/i" =>
1642                     Array(
1643                           Array(
1644                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1645                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1646                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1647                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1648                                 ),
1649                           Array(
1650                                 "\\1$secremoveimg\\2",
1651                                 "\\1$secremoveimg\\2",
1652                                 "\\1$secremoveimg\\2",
1653                                 "\\1$secremoveimg\\2"
1654                                 )
1655                         ),
1656                 "/^href|action/i" =>
1657                     Array(
1658                           Array(
1659                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1660                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1661                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1662                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1663                                 ),
1664                           Array(
1665                                 "\\1#\\2",
1666                                 "\\1#\\2",
1667                                 "\\1#\\2",
1668                                 "\\1#\\2"
1669                                 )
1670                         ),
1671                 "/^style/si" =>
1672                     Array(
1673                           Array(
1674                                 "/expression/si",
1675                                 "/binding/si",
1676                                 "/behaviou*r/si",
1677                                 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1678                                 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1679                                 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1680                                 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1681                                ),
1682                           Array(
1683                                 "idiocy",
1684                                 "idiocy",
1685                                 "idiocy",
1686                                 "url(\\1#\\2)",
1687                                 "url(\\1#\\2)",
1688                                 "url(\\1#\\2)",
1689                                 "url(\\1#\\2)"
1690                                )
1691                           )
1692                 )
1693         );
1694     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1695         $view_unsafe_images = false;
1696     }
1697     if (!$view_unsafe_images){
1698         /**
1699          * Remove any references to http/https if view_unsafe_images set
1700          * to false.
1701          */
1702          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1703                     '/^([\'\"])\s*https*:.*([\'\"])/si');
1704          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1705                     "\\1$secremoveimg\\2");
1706          array_push($bad_attvals{'/.*/'}{'/^style/si'}[0],
1707                     '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1708          array_push($bad_attvals{'/.*/'}{'/^style/si'}[1],
1709                     "url(\\1$secremoveimg\\2)");
1710     }
1711
1712     $add_attr_to_tag = Array(
1713                              "/^a$/si" => Array('target'=>'"_new"')
1714                              );
1715     $trusted = sq_sanitize($body,
1716                            $tag_list,
1717                            $rm_tags_with_content,
1718                            $self_closing_tags,
1719                            $force_tag_closing,
1720                            $rm_attnames,
1721                            $bad_attvals,
1722                            $add_attr_to_tag,
1723                            $message,
1724                            $id,
1725                            $mailbox
1726                            );
1727     if (preg_match("|$secremoveimg|si", $trusted)){
1728         $has_unsafe_images = true;
1729     }
1730     return $trusted;
1731 }
1732
1733 ?>