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     global $languages, $squirrelmail_language;
 324
 325     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
 326         $view_unsafe_images = false;
 327     }
 328
 329     $body = '';
 330     $urlmailbox = urlencode($mailbox);
 331     $body_message = getEntity($message, $ent_num);
 332     if (($body_message->header->type0 == 'text') ||
 333         ($body_message->header->type0 == 'rfc822')) {
 334         $body = mime_fetch_body ($imap_stream, $id, $ent_num);
 335         $body = decodeBody($body, $body_message->header->encoding);
 336
 337         if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 338             function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 339             if (mb_detect_encoding($body) != 'ASCII') {
 340                 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
 341             }
 342         }
 343         $hookResults = do_hook("message_body", $body);
 344         $body = $hookResults[1];
 345
 346         /* If there are other types that shouldn't be formatted, add
 347          * them here.
 348          */
 349
 350         if ($body_message->header->type1 == 'html') {
 351             if ($show_html_default <> 1) {
 352                 $entity_conv = array('&nbsp;' => ' ',
 353                                      '<p>'    => "\n",
 354                                      '<br>'   => "\n",
 355                                      '<P>'    => "\n",
 356                                      '<BR>'   => "\n",
 357                                      '&gt;'   => '>',
 358                                      '&lt;'   => '<');
 359                 $body = strtr($body, $entity_conv);
 360                 $body = strip_tags($body);
 361                 $body = trim($body);
 362                 translateText($body, $wrap_at,
 363                               $body_message->header->getParameter('charset'));
 364             } else {
 365                 $body = magicHTML($body, $id, $message, $mailbox);
 366             }
 367         } else {
 368             translateText($body, $wrap_at,
 369                           $body_message->header->getParameter('charset'));
 370         }
 371         $link = 'read_body.php?passed_id=' . $id . '&amp;ent_id='.$ent_num.
 372                 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
 373                 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 374         if (isset($passed_ent_id)) {
 375             $link .= '&amp;passed_ent_id='.$passed_ent_id;
 376         }
 377         if ($view_unsafe_images) {
 378             $text = _("Hide Unsafe Images");
 379         } else {
 380             if (isset($has_unsafe_images) && $has_unsafe_images) {
 381                 $link .= '&amp;view_unsafe_images=1';
 382                 $text = _("View Unsafe Images");
 383             } else {
 384                 $text = '';
 385             }
 386         }
 387         $body .= '<center><small><a href="'.$link.'">'.$text.
 388                  '</a></small></center><br>' . "\n";
 389     }
 390     return $body;
 391 }
 392
 393
 394 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 395     global $where, $what, $startMessage, $color;
 396     static $ShownHTML = 0;
 397
 398     $att_ar = $message->getAttachments($exclude_id);
 399
 400     if (!count($att_ar)) return '';
 401
 402     $attachments = '';
 403
 404     $urlMailbox = urlencode($mailbox);
 405
 406     foreach ($att_ar as $att) {
 407         $ent = $att->entity_id;
 408         $header = $att->header;
 409         $type0 = strtolower($header->type0);
 410         $type1 = strtolower($header->type1);
 411         $name = '';
 412         $links['download link']['text'] = _("download");
 413         $links['download link']['href'] = SM_PATH .
 414                 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 415         $ImageURL = '';
 416         if ($type0 =='message' && $type1 == 'rfc822') {
 417             $default_page = SM_PATH . 'src/read_body.php';
 418             $rfc822_header = $att->rfc822_header;
 419             $filename = $rfc822_header->subject;
 420             if (trim( $filename ) == '') {
 421                 $filename = 'untitled-[' . $ent . ']' ;
 422             }
 423             $from_o = $rfc822_header->from;
 424             if (is_object($from_o)) {
 425                 $from_name = $from_o->getAddress(false);
 426             } else {
 427                 $from_name = _("Unknown sender");
 428             }
 429             $from_name = decodeHeader(($from_name));
 430             $description = $from_name;
 431         } else {
 432             $default_page = SM_PATH . 'src/download.php';
 433             if (is_object($header->disposition)) {
 434                 $filename = $header->disposition->getProperty('filename');
 435                 if (trim($filename) == '') {
 436                     $name = decodeHeader($header->disposition->getProperty('name'));
 437                     if (trim($name) == '') {
 438                         $name = $header->getParameter('name');
 439                         if(trim($name) == '') {
 440                             if (trim( $header->id ) == '') {
 441                                 $filename = 'untitled-[' . $ent . ']' ;
 442                             } else {
 443                                 $filename = 'cid: ' . $header->id;
 444                             }
 445                         } else {
 446                             $filename = $name;
 447                         }
 448                     } else {
 449                         $filename = $name;
 450                     }
 451                 }
 452             } else {
 453                 $filename = $header->getParameter('name');
 454                 if (!trim($filename)) {
 455                     if (trim( $header->id ) == '') {
 456                         $filename = 'untitled-[' . $ent . ']' ;
 457                     } else {
 458                         $filename = 'cid: ' . $header->id;
 459                     }
 460                 }
 461             }
 462             if ($header->description) {
 463                 $description = decodeHeader($header->description);
 464             } else {
 465                 $description = '';
 466             }
 467         }
 468
 469         $display_filename = $filename;
 470         if (isset($passed_ent_id)) {
 471             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 472         } else {
 473             $passed_ent_id_link = '';
 474         }
 475         $defaultlink = $default_page . "?startMessage=$startMessage"
 476                      . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 477                      . '&amp;ent_id='.$ent.$passed_ent_id_link;
 478         if ($where && $what) {
 479            $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 480         }
 481         /* This executes the attachment hook with a specific MIME-type.
 482          * If that doesn't have results, it tries if there's a rule
 483          * for a more generic type.
 484          */
 485         $hookresults = do_hook("attachment $type0/$type1", $links,
 486                                $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 487                                $display_filename, $where, $what);
 488         if(count($hookresults[1]) <= 1) {
 489             $hookresults = do_hook("attachment $type0/*", $links,
 490                                    $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 491                                    $display_filename, $where, $what);
 492         }
 493
 494         $links = $hookresults[1];
 495         $defaultlink = $hookresults[6];
 496
 497         $attachments .= '<TR><TD>' .
 498                         '<A HREF="'.$defaultlink.'">'.decodeHeader($display_filename).'</A>&nbsp;</TD>' .
 499                         '<TD><SMALL><b>' . show_readable_size($header->size) .
 500                         '</b>&nbsp;&nbsp;</small></TD>' .
 501                         "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
 502                         '<TD><SMALL>';
 503         $attachments .= '<b>' . $description . '</b>';
 504         $attachments .= '</SMALL></TD><TD><SMALL>&nbsp;';
 505
 506         $skipspaces = 1;
 507         foreach ($links as $val) {
 508             if ($skipspaces) {
 509                 $skipspaces = 0;
 510             } else {
 511                 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
 512             }
 513             $attachments .= '<a href="' . $val['href'] . '">' .  $val['text'] . '</a>';
 514         }
 515         unset($links);
 516         $attachments .= "</TD></TR>\n";
 517     }
 518     return $attachments;
 519 }
 520
 521 /* This function decodes the body depending on the encoding type. */
 522 function decodeBody($body, $encoding) {
 523     global $show_html_default;
 524
 525     $body = str_replace("\r\n", "\n", $body);
 526     $encoding = strtolower($encoding);
 527
 528     if ($encoding == 'quoted-printable' ||
 529         $encoding == 'quoted_printable') {
 530         $body = quoted_printable_decode($body);
 531
 532         while (ereg("=\n", $body)) {
 533             $body = ereg_replace ("=\n", '', $body);
 534         }
 535
 536     } else if ($encoding == 'base64') {
 537         $body = base64_decode($body);
 538     }
 539
 540     // All other encodings are returned raw.
 541     return $body;
 542 }
 543
 544 /*
 545  * This functions decode strings that is encoded according to
 546  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 547  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 548  */
 549 function decodeHeader ($string, $utfencode=true,$htmlsave=true) {
 550     global $languages, $squirrelmail_language;
 551     if (is_array($string)) {
 552         $string = implode("\n", $string);
 553     }
 554
 555     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 556         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 557         $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
 558         // Do we need to return at this point?
 559         // return $string;
 560     }
 561     $i = 0;
 562     $iLastMatch = -2;
 563     $encoded = false;
 564
 565     $aString = explode(' ',$string);
 566     $ret = '';
 567     foreach ($aString as $chunk) {
 568         if ($encoded && !$chunk) {
 569             continue;
 570         } elseif (!$chunk) {
 571             $ret .= ' ';
 572             continue;
 573         }
 574         $encoded = false;
 575         /* if encoded words are not separated by a linear-space-white we still catch them */
 576         $j = $i-1;
 577         while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 578             /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
 579             if ($iLastMatch !== $j) {
 580                 if ($htmlsave) {
 581                     $ret .= '&nbsp;';
 582                 } else {
 583                     $ret .= ' ';
 584                 }
 585             }
 586             $iLastMatch = $i;
 587             $j = $i;
 588             $ret .= $res[1];
 589             $encoding = ucfirst($res[3]);
 590             switch ($encoding)
 591             {
 592             case 'B':
 593                 $replace = base64_decode($res[4]);
 594                 $ret .= charset_decode($res[2],$replace);
 595                 break;
 596             case 'Q':
 597                 $replace = str_replace('_', ' ', $res[4]);
 598                 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 599                                     $replace);
 600                 /* Only encode into entities by default. Some places
 601                  * don't need the encoding, like the compose form.
 602                  */
 603                 if ($utfencode) {
 604                     $replace = charset_decode($res[2], $replace);
 605                 } else {
 606                     if ($htmlsave) {
 607                         $replace = htmlspecialchars($replace);
 608                     }
 609                 }
 610                 $ret .= $replace;
 611                 break;
 612             default:
 613                 break;
 614             }
 615             $chunk = $res[5];
 616             $encoded = true;
 617         }
 618         if (!$encoded) {
 619             if ($htmlsave) {
 620                 $ret .= '&nbsp;';
 621             } else {
 622                 $ret .= ' ';
 623             }
 624         }
 625
 626         if (!$encoded && $htmlsave) {
 627             $ret .= htmlspecialchars($chunk);
 628         } else {
 629             $ret .= $chunk;
 630         }
 631         ++$i;
 632     }
 633     /* remove the first added space */
 634     if ($ret) {
 635         if ($htmlsave) {
 636             $ret = substr($ret,6);
 637         } else {
 638             $ret = substr($ret,1);
 639         }
 640     }
 641
 642     return $ret;
 643 }
 644
 645 /*
 646  * Encode a string according to RFC 1522 for use in headers if it
 647  * contains 8-bit characters or anything that looks like it should
 648  * be encoded.
 649  */
 650 function encodeHeader ($string) {
 651     global $default_charset, $languages, $squirrelmail_language;
 652
 653     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 654         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 655         return  $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
 656     }
 657     if (strtolower($default_charset) == 'iso-8859-1') {
 658         $string = str_replace("\240",' ',$string);
 659     }
 660
 661     // Encode only if the string contains 8-bit characters or =?
 662     $j = strlen($string);
 663     $max_l = 75 - strlen($default_charset) - 7;
 664     $aRet = array();
 665     $ret = '';
 666     $iEncStart = $enc_init = false;
 667     $cur_l = $iOffset = 0;
 668     for($i = 0; $i < $j; ++$i) {
 669         switch($string{$i})
 670         {
 671         case '=':
 672         case '<':
 673         case '>':
 674         case ',':
 675         case '?':
 676         case '_':
 677             if ($iEncStart === false) {
 678                 $iEncStart = $i;
 679             }
 680             $cur_l+=3;
 681             if ($cur_l > ($max_l-2)) {
 682                 /* if there is an stringpart that doesn't need encoding, add it */
 683                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 684                 $aRet[] = "=?$default_charset?Q?$ret?=";
 685                 $iOffset = $i;
 686                 $cur_l = 0;
 687                 $ret = '';
 688                 $iEncStart = false;
 689             } else {
 690                 $ret .= sprintf("=%02X",ord($string{$i}));
 691             }
 692             break;
 693         case '(':
 694         case ')':
 695             if ($iEncStart !== false) {
 696                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 697                 $aRet[] = "=?$default_charset?Q?$ret?=";
 698                 $iOffset = $i;
 699                 $cur_l = 0;
 700                 $ret = '';
 701                 $iEncStart = false;
 702             }
 703             break;
 704         case ' ':
 705             if ($iEncStart !== false) {
 706                 $cur_l++;
 707                 if ($cur_l > $max_l) {
 708                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 709                     $aRet[] = "=?$default_charset?Q?$ret?=";
 710                     $iOffset = $i;
 711                     $cur_l = 0;
 712                     $ret = '';
 713                     $iEncStart = false;
 714                 } else {
 715                     $ret .= '_';
 716                 }
 717             }
 718             break;
 719         default:
 720             $k = ord($string{$i});
 721             if ($k > 126) {
 722                 if ($iEncStart === false) {
 723                     $iEncStart = $i;
 724                 }
 725                 $cur_l += 3;
 726                 /* first we add the encoded string that reached it's max size */
 727                 if ($cur_l > ($max_l-2)) {
 728                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 729                     $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
 730                     $cur_l = 3;
 731                     $ret = '';
 732                     $iOffset = $i;
 733                     $iEncStart = $i;
 734                 }
 735                 $enc_init = true;
 736                 $ret .= sprintf("=%02X", $k);
 737             } else {
 738                 if ($iEncStart !== false) {
 739                     $cur_l++;
 740                     if ($cur_l > $max_l) {
 741                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 742                         $aRet[] = "=?$default_charset?Q?$ret?=";
 743                         $iEncStart = false;
 744                         $iOffset = $i;
 745                         $cur_l = 0;
 746                         $ret = '';
 747                     } else {
 748                         $ret .= $string{$i};
 749                     }
 750                 }
 751             }
 752             break;
 753         }
 754     }
 755
 756     if ($enc_init) {
 757         if ($iEncStart !== false) {
 758             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 759             $aRet[] = "=?$default_charset?Q?$ret?=";
 760         } else {
 761             $aRet[] = substr($string,$iOffset);
 762         }
 763         $string = implode('',$aRet);
 764     }
 765     return $string;
 766 }
 767
 768 /* This function trys to locate the entity_id of a specific mime element */
 769 function find_ent_id($id, $message) {
 770     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
 771         if ($message->entities[$i]->header->type0 == 'multipart')  {
 772             $ret = find_ent_id($id, $message->entities[$i]);
 773         } else {
 774             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
 775 //                if (sq_check_save_extension($message->entities[$i])) {
 776                     return $message->entities[$i]->entity_id;
 777 //                }
 778             }
 779         }
 780     }
 781     return $ret;
 782 }
 783
 784 function sq_check_save_extension($message) {
 785     $filename = $message->getFilename();
 786     $ext = substr($filename, strrpos($filename,'.')+1);
 787     $save_extensions = array('jpg','jpeg','gif','png','bmp');
 788     return in_array($ext, $save_extensions);
 789 }
 790
 791
 792 /**
 793  ** HTMLFILTER ROUTINES
 794  */
 795
 796 /**
 797  * This function returns the final tag out of the tag name, an array
 798  * of attributes, and the type of the tag. This function is called by
 799  * sq_sanitize internally.
 800  *
 801  * @param  $tagname  the name of the tag.
 802  * @param  $attary   the array of attributes and their values
 803  * @param  $tagtype  The type of the tag (see in comments).
 804  * @return           a string with the final tag representation.
 805  */
 806 function sq_tagprint($tagname, $attary, $tagtype){
 807     $me = 'sq_tagprint';
 808
 809     if ($tagtype == 2){
 810         $fulltag = '</' . $tagname . '>';
 811     } else {
 812         $fulltag = '<' . $tagname;
 813         if (is_array($attary) && sizeof($attary)){
 814             $atts = Array();
 815             while (list($attname, $attvalue) = each($attary)){
 816                 array_push($atts, "$attname=$attvalue");
 817             }
 818             $fulltag .= ' ' . join(" ", $atts);
 819         }
 820         if ($tagtype == 3){
 821             $fulltag .= ' /';
 822         }
 823         $fulltag .= '>';
 824     }
 825     return $fulltag;
 826 }
 827
 828 /**
 829  * A small helper function to use with array_walk. Modifies a by-ref
 830  * value and makes it lowercase.
 831  *
 832  * @param  $val a value passed by-ref.
 833  * @return      void since it modifies a by-ref value.
 834  */
 835 function sq_casenormalize(&$val){
 836     $val = strtolower($val);
 837 }
 838
 839 /**
 840  * This function skips any whitespace from the current position within
 841  * a string and to the next non-whitespace value.
 842  *
 843  * @param  $body   the string
 844  * @param  $offset the offset within the string where we should start
 845  *                 looking for the next non-whitespace character.
 846  * @return         the location within the $body where the next
 847  *                 non-whitespace char is located.
 848  */
 849 function sq_skipspace($body, $offset){
 850     $me = 'sq_skipspace';
 851     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
 852     if (sizeof($matches{1})){
 853         $count = strlen($matches{1});
 854         $offset += $count;
 855     }
 856     return $offset;
 857 }
 858
 859 /**
 860  * This function looks for the next character within a string.  It's
 861  * really just a glorified "strpos", except it catches if failures
 862  * nicely.
 863  *
 864  * @param  $body   The string to look for needle in.
 865  * @param  $offset Start looking from this position.
 866  * @param  $needle The character/string to look for.
 867  * @return         location of the next occurance of the needle, or
 868  *                 strlen($body) if needle wasn't found.
 869  */
 870 function sq_findnxstr($body, $offset, $needle){
 871     $me  = 'sq_findnxstr';
 872     $pos = strpos($body, $needle, $offset);
 873     if ($pos === FALSE){
 874         $pos = strlen($body);
 875     }
 876     return $pos;
 877 }
 878
 879 /**
 880  * This function takes a PCRE-style regexp and tries to match it
 881  * within the string.
 882  *
 883  * @param  $body   The string to look for needle in.
 884  * @param  $offset Start looking from here.
 885  * @param  $reg    A PCRE-style regex to match.
 886  * @return         Returns a false if no matches found, or an array
 887  *                 with the following members:
 888  *                 - integer with the location of the match within $body
 889  *                 - string with whatever content between offset and the match
 890  *                 - string with whatever it is we matched
 891  */
 892 function sq_findnxreg($body, $offset, $reg){
 893     $me = 'sq_findnxreg';
 894     $matches = Array();
 895     $retarr = Array();
 896     preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
 897     if (!$matches{0}){
 898         $retarr = false;
 899     } else {
 900         $retarr{0} = $offset + strlen($matches{1});
 901         $retarr{1} = $matches{1};
 902         $retarr{2} = $matches{2};
 903     }
 904     return $retarr;
 905 }
 906
 907 /**
 908  * This function looks for the next tag.
 909  *
 910  * @param  $body   String where to look for the next tag.
 911  * @param  $offset Start looking from here.
 912  * @return         false if no more tags exist in the body, or
 913  *                 an array with the following members:
 914  *                 - string with the name of the tag
 915  *                 - array with attributes and their values
 916  *                 - integer with tag type (1, 2, or 3)
 917  *                 - integer where the tag starts (starting "<")
 918  *                 - integer where the tag ends (ending ">")
 919  *                 first three members will be false, if the tag is invalid.
 920  */
 921 function sq_getnxtag($body, $offset){
 922     $me = 'sq_getnxtag';
 923     if ($offset > strlen($body)){
 924         return false;
 925     }
 926     $lt = sq_findnxstr($body, $offset, "<");
 927     if ($lt == strlen($body)){
 928         return false;
 929     }
 930     /**
 931      * We are here:
 932      * blah blah <tag attribute="value">
 933      * \---------^
 934      */
 935     $pos = sq_skipspace($body, $lt+1);
 936     if ($pos >= strlen($body)){
 937         return Array(false, false, false, $lt, strlen($body));
 938     }
 939     /**
 940      * There are 3 kinds of tags:
 941      * 1. Opening tag, e.g.:
 942      *    <a href="blah">
 943      * 2. Closing tag, e.g.:
 944      *    </a>
 945      * 3. XHTML-style content-less tag, e.g.:
 946      *    <img src="blah"/>
 947      */
 948     $tagtype = false;
 949     switch (substr($body, $pos, 1)){
 950         case '/':
 951             $tagtype = 2;
 952             $pos++;
 953             break;
 954         case '!':
 955             /**
 956              * A comment or an SGML declaration.
 957              */
 958             if (substr($body, $pos+1, 2) == "--"){
 959                 $gt = strpos($body, "-->", $pos);
 960                 if ($gt === false){
 961                     $gt = strlen($body);
 962                 } else {
 963                     $gt += 2;
 964                 }
 965                 return Array(false, false, false, $lt, $gt);
 966             } else {
 967                 $gt = sq_findnxstr($body, $pos, ">");
 968                 return Array(false, false, false, $lt, $gt);
 969             }
 970             break;
 971         default:
 972             /**
 973              * Assume tagtype 1 for now. If it's type 3, we'll switch values
 974              * later.
 975              */
 976             $tagtype = 1;
 977             break;
 978     }
 979
 980     $tag_start = $pos;
 981     $tagname = '';
 982     /**
 983      * Look for next [\W-_], which will indicate the end of the tag name.
 984      */
 985     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
 986     if ($regary == false){
 987         return Array(false, false, false, $lt, strlen($body));
 988     }
 989     list($pos, $tagname, $match) = $regary;
 990     $tagname = strtolower($tagname);
 991
 992     /**
 993      * $match can be either of these:
 994      * '>'  indicating the end of the tag entirely.
 995      * '\s' indicating the end of the tag name.
 996      * '/'  indicating that this is type-3 xhtml tag.
 997      *
 998      * Whatever else we find there indicates an invalid tag.
 999      */
1000     switch ($match){
1001         case '/':
1002             /**
1003              * This is an xhtml-style tag with a closing / at the
1004              * end, like so: <img src="blah"/>. Check if it's followed
1005              * by the closing bracket. If not, then this tag is invalid
1006              */
1007             if (substr($body, $pos, 2) == "/>"){
1008                 $pos++;
1009                 $tagtype = 3;
1010             } else {
1011                 $gt = sq_findnxstr($body, $pos, ">");
1012                 $retary = Array(false, false, false, $lt, $gt);
1013                 return $retary;
1014             }
1015         case '>':
1016             return Array($tagname, false, $tagtype, $lt, $pos);
1017             break;
1018         default:
1019             /**
1020              * Check if it's whitespace
1021              */
1022             if (!preg_match('/\s/', $match)){
1023                 /**
1024                  * This is an invalid tag! Look for the next closing ">".
1025                  */
1026                 $gt = sq_findnxstr($body, $offset, ">");
1027                 return Array(false, false, false, $lt, $gt);
1028             }
1029             break;
1030     }
1031
1032     /**
1033      * At this point we're here:
1034      * <tagname  attribute='blah'>
1035      * \-------^
1036      *
1037      * At this point we loop in order to find all attributes.
1038      */
1039     $attname = '';
1040     $atttype = false;
1041     $attary = Array();
1042
1043     while ($pos <= strlen($body)){
1044         $pos = sq_skipspace($body, $pos);
1045         if ($pos == strlen($body)){
1046             /**
1047              * Non-closed tag.
1048              */
1049             return Array(false, false, false, $lt, $pos);
1050         }
1051         /**
1052          * See if we arrived at a ">" or "/>", which means that we reached
1053          * the end of the tag.
1054          */
1055         $matches = Array();
1056         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1057             /**
1058              * Yep. So we did.
1059              */
1060             $pos += strlen($matches{1});
1061             if ($matches{2} == "/>"){
1062                 $tagtype = 3;
1063                 $pos++;
1064             }
1065             return Array($tagname, $attary, $tagtype, $lt, $pos);
1066         }
1067
1068         /**
1069          * There are several types of attributes, with optional
1070          * [:space:] between members.
1071          * Type 1:
1072          *   attrname[:space:]=[:space:]'CDATA'
1073          * Type 2:
1074          *   attrname[:space:]=[:space:]"CDATA"
1075          * Type 3:
1076          *   attr[:space:]=[:space:]CDATA
1077          * Type 4:
1078          *   attrname
1079          *
1080          * We leave types 1 and 2 the same, type 3 we check for
1081          * '"' and convert to "&quot" if needed, then wrap in
1082          * double quotes. Type 4 we convert into:
1083          * attrname="yes".
1084          */
1085         $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1086         if ($regary == false){
1087             /**
1088              * Looks like body ended before the end of tag.
1089              */
1090             return Array(false, false, false, $lt, strlen($body));
1091         }
1092         list($pos, $attname, $match) = $regary;
1093         $attname = strtolower($attname);
1094         /**
1095          * We arrived at the end of attribute name. Several things possible
1096          * here:
1097          * '>'  means the end of the tag and this is attribute type 4
1098          * '/'  if followed by '>' means the same thing as above
1099          * '\s' means a lot of things -- look what it's followed by.
1100          *      anything else means the attribute is invalid.
1101          */
1102         switch($match){
1103             case '/':
1104                 /**
1105                  * This is an xhtml-style tag with a closing / at the
1106                  * end, like so: <img src="blah"/>. Check if it's followed
1107                  * by the closing bracket. If not, then this tag is invalid
1108                  */
1109                 if (substr($body, $pos, 2) == "/>"){
1110                     $pos++;
1111                     $tagtype = 3;
1112                 } else {
1113                     $gt = sq_findnxstr($body, $pos, ">");
1114                     $retary = Array(false, false, false, $lt, $gt);
1115                     return $retary;
1116                 }
1117             case '>':
1118                 $attary{$attname} = '"yes"';
1119                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1120                 break;
1121             default:
1122                 /**
1123                  * Skip whitespace and see what we arrive at.
1124                  */
1125                 $pos = sq_skipspace($body, $pos);
1126                 $char = substr($body, $pos, 1);
1127                 /**
1128                  * Two things are valid here:
1129                  * '=' means this is attribute type 1 2 or 3.
1130                  * \w means this was attribute type 4.
1131                  * anything else we ignore and re-loop. End of tag and
1132                  * invalid stuff will be caught by our checks at the beginning
1133                  * of the loop.
1134                  */
1135                 if ($char == "="){
1136                     $pos++;
1137                     $pos = sq_skipspace($body, $pos);
1138                     /**
1139                      * Here are 3 possibilities:
1140                      * "'"  attribute type 1
1141                      * '"'  attribute type 2
1142                      * everything else is the content of tag type 3
1143                      */
1144                     $quot = substr($body, $pos, 1);
1145                     if ($quot == "'"){
1146                         $regary = sq_findnxreg($body, $pos+1, "\'");
1147                         if ($regary == false){
1148                             return Array(false, false, false, $lt, strlen($body));
1149                         }
1150                         list($pos, $attval, $match) = $regary;
1151                         $pos++;
1152                         $attary{$attname} = "'" . $attval . "'";
1153                     } else if ($quot == '"'){
1154                         $regary = sq_findnxreg($body, $pos+1, '\"');
1155                         if ($regary == false){
1156                             return Array(false, false, false, $lt, strlen($body));
1157                         }
1158                         list($pos, $attval, $match) = $regary;
1159                         $pos++;
1160                         $attary{$attname} = '"' . $attval . '"';
1161                     } else {
1162                         /**
1163                          * These are hateful. Look for \s, or >.
1164                          */
1165                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1166                         if ($regary == false){
1167                             return Array(false, false, false, $lt, strlen($body));
1168                         }
1169                         list($pos, $attval, $match) = $regary;
1170                         /**
1171                          * If it's ">" it will be caught at the top.
1172                          */
1173                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1174                         $attary{$attname} = '"' . $attval . '"';
1175                     }
1176                 } else if (preg_match("|[\w/>]|", $char)) {
1177                     /**
1178                      * That was attribute type 4.
1179                      */
1180                     $attary{$attname} = '"yes"';
1181                 } else {
1182                     /**
1183                      * An illegal character. Find next '>' and return.
1184                      */
1185                     $gt = sq_findnxstr($body, $pos, ">");
1186                     return Array(false, false, false, $lt, $gt);
1187                 }
1188                 break;
1189         }
1190     }
1191     /**
1192      * The fact that we got here indicates that the tag end was never
1193      * found. Return invalid tag indication so it gets stripped.
1194      */
1195     return Array(false, false, false, $lt, strlen($body));
1196 }
1197
1198 /**
1199  * This function checks attribute values for entity-encoded values
1200  * and returns them translated into 8-bit strings so we can run
1201  * checks on them.
1202  *
1203  * @param  $attvalue A string to run entity check against.
1204  * @return           Translated value.
1205  */
1206 function sq_deent($attvalue){
1207     $me = 'sq_deent';
1208     /**
1209      * See if we have to run the checks first. All entities must start
1210      * with "&".
1211      */
1212     if (strpos($attvalue, "&") === false){
1213         return $attvalue;
1214     }
1215     /**
1216      * Check named entities first.
1217      */
1218     $trans = get_html_translation_table(HTML_ENTITIES);
1219     /**
1220      * Leave &quot; in, as it can mess us up.
1221      */
1222     $trans = array_flip($trans);
1223     unset($trans{"&quot;"});
1224     while (list($ent, $val) = each($trans)){
1225         $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1226     }
1227     /**
1228      * Now translate numbered entities from 1 to 255 if needed.
1229      */
1230     if (strpos($attvalue, "#") !== false){
1231         $omit = Array(34, 39);
1232         for ($asc=1; $asc<256; $asc++){
1233             if (!in_array($asc, $omit)){
1234                 $chr = chr($asc);
1235                 $attvalue = preg_replace("/\&#0*$asc;*(\D)/si", "$chr\\1",
1236                                          $attvalue);
1237                 $attvalue = preg_replace("/\&#x0*".dechex($asc).";*(\W)/si",
1238                                          "$chr\\1", $attvalue);
1239             }
1240         }
1241     }
1242     return $attvalue;
1243 }
1244
1245 /**
1246  * This function runs various checks against the attributes.
1247  *
1248  * @param  $tagname         String with the name of the tag.
1249  * @param  $attary          Array with all tag attributes.
1250  * @param  $rm_attnames     See description for sq_sanitize
1251  * @param  $bad_attvals     See description for sq_sanitize
1252  * @param  $add_attr_to_tag See description for sq_sanitize
1253  * @param  $message         message object
1254  * @param  $id              message id
1255  * @return                  Array with modified attributes.
1256  */
1257 function sq_fixatts($tagname,
1258                     $attary,
1259                     $rm_attnames,
1260                     $bad_attvals,
1261                     $add_attr_to_tag,
1262                     $message,
1263                     $id,
1264                     $mailbox
1265                     ){
1266     $me = 'sq_fixatts';
1267     while (list($attname, $attvalue) = each($attary)){
1268         /**
1269          * See if this attribute should be removed.
1270          */
1271         foreach ($rm_attnames as $matchtag=>$matchattrs){
1272             if (preg_match($matchtag, $tagname)){
1273                 foreach ($matchattrs as $matchattr){
1274                     if (preg_match($matchattr, $attname)){
1275                         unset($attary{$attname});
1276                         continue;
1277                     }
1278                 }
1279             }
1280         }
1281         /**
1282          * Remove any entities.
1283          */
1284         $attvalue = sq_deent($attvalue);
1285
1286         /**
1287          * Now let's run checks on the attvalues.
1288          * I don't expect anyone to comprehend this. If you do,
1289          * get in touch with me so I can drive to where you live and
1290          * shake your hand personally. :)
1291          */
1292         foreach ($bad_attvals as $matchtag=>$matchattrs){
1293             if (preg_match($matchtag, $tagname)){
1294                 foreach ($matchattrs as $matchattr=>$valary){
1295                     if (preg_match($matchattr, $attname)){
1296                         /**
1297                          * There are two arrays in valary.
1298                          * First is matches.
1299                          * Second one is replacements
1300                          */
1301                         list($valmatch, $valrepl) = $valary;
1302                         $newvalue =
1303                             preg_replace($valmatch, $valrepl, $attvalue);
1304                         if ($newvalue != $attvalue){
1305                             $attary{$attname} = $newvalue;
1306                         }
1307                     }
1308                 }
1309             }
1310         }
1311         /**
1312          * Turn cid: urls into http-friendly ones.
1313          */
1314         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1315             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1316         }
1317     }
1318     /**
1319      * See if we need to append any attributes to this tag.
1320      */
1321     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1322         if (preg_match($matchtag, $tagname)){
1323             $attary = array_merge($attary, $addattary);
1324         }
1325     }
1326     return $attary;
1327 }
1328
1329 /**
1330  * This function edits the style definition to make them friendly and
1331  * usable in squirrelmail.
1332  *
1333  * @param  $message  the message object
1334  * @param  $id       the message id
1335  * @param  $content  a string with whatever is between <style> and </style>
1336  * @return           a string with edited content.
1337  */
1338 function sq_fixstyle($message, $id, $content){
1339     global $view_unsafe_images;
1340     $me = 'sq_fixstyle';
1341     /**
1342      * First look for general BODY style declaration, which would be
1343      * like so:
1344      * body {background: blah-blah}
1345      * and change it to .bodyclass so we can just assign it to a <div>
1346      */
1347     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1348     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1349     /**
1350      * Fix url('blah') declarations.
1351      */
1352     $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1353                             "url(\\1$secremoveimg\\2)", $content);
1354     /**
1355      * Fix url('https*://.*) declarations but only if $view_unsafe_images
1356      * is false.
1357      */
1358     if (!$view_unsafe_images){
1359         $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1360                                 "url(\\1$secremoveimg\\2)", $content);
1361     }
1362
1363     /**
1364      * Fix urls that refer to cid:
1365      */
1366     while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1367                       $matches)){
1368         $cidurl = $matches{1};
1369         $httpurl = sq_cid2http($message, $id, $cidurl);
1370         $content = preg_replace("|url\($cidurl\)|si",
1371                                 "url($httpurl)", $content);
1372     }
1373
1374     /**
1375      * Fix stupid css declarations which lead to vulnerabilities
1376      * in IE.
1377      */
1378     $match   = Array('/expression/i',
1379                      '/behaviou*r/i',
1380                      '/binding/i');
1381     $replace = Array('idiocy', 'idiocy', 'idiocy');
1382     $content = preg_replace($match, $replace, $content);
1383     return $content;
1384 }
1385
1386 /**
1387  * This function converts cid: url's into the ones that can be viewed in
1388  * the browser.
1389  *
1390  * @param  $message  the message object
1391  * @param  $id       the message id
1392  * @param  $cidurl   the cid: url.
1393  * @return           a string with a http-friendly url
1394  */
1395 function sq_cid2http($message, $id, $cidurl, $mailbox){
1396     /**
1397      * Get rid of quotes.
1398      */
1399     $quotchar = substr($cidurl, 0, 1);
1400     $cidurl = str_replace($quotchar, "", $cidurl);
1401     $cidurl = substr(trim($cidurl), 4);
1402     $linkurl = find_ent_id($cidurl, $message);
1403     /* in case of non-save cid links $httpurl should be replaced by a sort of
1404        unsave link image */
1405     $httpurl = '';
1406     if ($linkurl) {
1407         $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1408                    "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1409                    '&amp;ent_id=' . $linkurl . $quotchar;
1410     }
1411     return $httpurl;
1412 }
1413
1414 /**
1415  * This function changes the <body> tag into a <div> tag since we
1416  * can't really have a body-within-body.
1417  *
1418  * @param  $attary  an array of attributes and values of <body>
1419  * @return          a modified array of attributes to be set for <div>
1420  */
1421 function sq_body2div($attary){
1422     $me = 'sq_body2div';
1423     $divattary = Array('class' => "'bodyclass'");
1424     $bgcolor = '#ffffff';
1425     $text = '#000000';
1426     $styledef = '';
1427     if (is_array($attary) && sizeof($attary) > 0){
1428         foreach ($attary as $attname=>$attvalue){
1429             $quotchar = substr($attvalue, 0, 1);
1430             $attvalue = str_replace($quotchar, "", $attvalue);
1431             switch ($attname){
1432                 case 'background':
1433                     $styledef .= "background-image: url('$attvalue'); ";
1434                     break;
1435                 case 'bgcolor':
1436                     $styledef .= "background-color: $attvalue; ";
1437                     break;
1438                 case 'text':
1439                     $styledef .= "color: $attvalue; ";
1440                     break;
1441             }
1442         }
1443         if (strlen($styledef) > 0){
1444             $divattary{"style"} = "\"$styledef\"";
1445         }
1446     }
1447     return $divattary;
1448 }
1449
1450 /**
1451  * This is the main function and the one you should actually be calling.
1452  * There are several variables you should be aware of an which need
1453  * special description.
1454  *
1455  * Since the description is quite lengthy, see it here:
1456  * http://www.mricon.com/html/phpfilter.html
1457  *
1458  * @param $body                 the string with HTML you wish to filter
1459  * @param $tag_list             see description above
1460  * @param $rm_tags_with_content see description above
1461  * @param $self_closing_tags    see description above
1462  * @param $force_tag_closing    see description above
1463  * @param $rm_attnames          see description above
1464  * @param $bad_attvals          see description above
1465  * @param $add_attr_to_tag      see description above
1466  * @param $message              message object
1467  * @param $id                   message id
1468  * @return                      sanitized html safe to show on your pages.
1469  */
1470 function sq_sanitize($body,
1471                      $tag_list,
1472                      $rm_tags_with_content,
1473                      $self_closing_tags,
1474                      $force_tag_closing,
1475                      $rm_attnames,
1476                      $bad_attvals,
1477                      $add_attr_to_tag,
1478                      $message,
1479                      $id,
1480                      $mailbox
1481                      ){
1482     $me = 'sq_sanitize';
1483     /**
1484      * Normalize rm_tags and rm_tags_with_content.
1485      */
1486     @array_walk($rm_tags, 'sq_casenormalize');
1487     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1488     @array_walk($self_closing_tags, 'sq_casenormalize');
1489     /**
1490      * See if tag_list is of tags to remove or tags to allow.
1491      * false  means remove these tags
1492      * true   means allow these tags
1493      */
1494     $rm_tags = array_shift($tag_list);
1495     $curpos = 0;
1496     $open_tags = Array();
1497     $trusted = "<!-- begin sanitized html -->\n";
1498     $skip_content = false;
1499     /**
1500      * Take care of netscape's stupid javascript entities like
1501      * &{alert('boo')};
1502      */
1503     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1504
1505     while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1506         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1507         $free_content = substr($body, $curpos, $lt-$curpos);
1508         /**
1509          * Take care of <style>
1510          */
1511         if ($tagname == "style" && $tagtype == 2){
1512             /**
1513              * This is a closing </style>. Edit the
1514              * content before we apply it.
1515              */
1516             $free_content = sq_fixstyle($message, $id, $free_content);
1517         }
1518         if ($skip_content == false){
1519             $trusted .= $free_content;
1520         }
1521         if ($tagname != FALSE){
1522             if ($tagtype == 2){
1523                 if ($skip_content == $tagname){
1524                     /**
1525                      * Got to the end of tag we needed to remove.
1526                      */
1527                     $tagname = false;
1528                     $skip_content = false;
1529                 } else {
1530                     if ($skip_content == false){
1531                         if ($tagname == "body"){
1532                             $tagname = "div";
1533                         } else {
1534                             if (isset($open_tags{$tagname}) &&
1535                                 $open_tags{$tagname} > 0){
1536                                 $open_tags{$tagname}--;
1537                             } else {
1538                                 $tagname = false;
1539                             }
1540                         }
1541                     }
1542                 }
1543             } else {
1544                 /**
1545                  * $rm_tags_with_content
1546                  */
1547                 if ($skip_content == false){
1548                     /**
1549                      * See if this is a self-closing type and change
1550                      * tagtype appropriately.
1551                      */
1552                     if ($tagtype == 1
1553                         && in_array($tagname, $self_closing_tags)){
1554                         $tagtype=3;
1555                     }
1556                     /**
1557                      * See if we should skip this tag and any content
1558                      * inside it.
1559                      */
1560                     if ($tagtype == 1 &&
1561                         in_array($tagname, $rm_tags_with_content)){
1562                         $skip_content = $tagname;
1563                     } else {
1564                         if (($rm_tags == false
1565                              && in_array($tagname, $tag_list)) ||
1566                             ($rm_tags == true &&
1567                              !in_array($tagname, $tag_list))){
1568                             $tagname = false;
1569                         } else {
1570                             if ($tagtype == 1){
1571                                 if (isset($open_tags{$tagname})){
1572                                     $open_tags{$tagname}++;
1573                                 } else {
1574                                     $open_tags{$tagname}=1;
1575                                 }
1576                             }
1577                             /**
1578                              * This is where we run other checks.
1579                              */
1580                             if (is_array($attary) && sizeof($attary) > 0){
1581                                 $attary = sq_fixatts($tagname,
1582                                                      $attary,
1583                                                      $rm_attnames,
1584                                                      $bad_attvals,
1585                                                      $add_attr_to_tag,
1586                                                      $message,
1587                                                      $id,
1588                                                      $mailbox
1589                                                      );
1590                             }
1591                             /**
1592                              * Convert body into div.
1593                              */
1594                             if ($tagname == "body"){
1595                                 $tagname = "div";
1596                                 $attary = sq_body2div($attary, $message, $id);
1597                             }
1598                         }
1599                     }
1600                 }
1601             }
1602             if ($tagname != false && $skip_content == false){
1603                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1604             }
1605         }
1606         $curpos = $gt+1;
1607     }
1608     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1609     if ($force_tag_closing == true){
1610         foreach ($open_tags as $tagname=>$opentimes){
1611             while ($opentimes > 0){
1612                 $trusted .= '</' . $tagname . '>';
1613                 $opentimes--;
1614             }
1615         }
1616         $trusted .= "\n";
1617     }
1618     $trusted .= "<!-- end sanitized html -->\n";
1619     return $trusted;
1620 }
1621
1622 /**
1623  * This is a wrapper function to call html sanitizing routines.
1624  *
1625  * @param  $body  the body of the message
1626  * @param  $id    the id of the message
1627  * @return        a string with html safe to display in the browser.
1628  */
1629 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1630     global $attachment_common_show_images, $view_unsafe_images,
1631            $has_unsafe_images;
1632     /**
1633      * Don't display attached images in HTML mode.
1634      */
1635     $attachment_common_show_images = false;
1636     $tag_list = Array(
1637                       false,
1638                       "object",
1639                       "meta",
1640                       "html",
1641                       "head",
1642                       "base",
1643                       "link",
1644                       "frame",
1645                       "iframe"
1646                       );
1647
1648     $rm_tags_with_content = Array(
1649                                   "script",
1650                                   "applet",
1651                                   "embed",
1652                                   "title"
1653                                   );
1654
1655     $self_closing_tags =  Array(
1656                                 "img",
1657                                 "br",
1658                                 "hr",
1659                                 "input"
1660                                 );
1661
1662     $force_tag_closing = false;
1663
1664     $rm_attnames = Array(
1665                          "/.*/" =>
1666                          Array(
1667                                "/target/i",
1668                                "/^on.*/i",
1669                                "/^dynsrc/i",
1670                                "/^data.*/i",
1671                                "/^lowsrc.*/i"
1672                                )
1673                          );
1674
1675     $secremoveimg = "../images/" . _("sec_remove_eng.png");
1676     $bad_attvals = Array(
1677         "/.*/" =>
1678             Array(
1679                 "/^src|background/i" =>
1680                     Array(
1681                           Array(
1682                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1683                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1684                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1685                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1686                                 ),
1687                           Array(
1688                                 "\\1$secremoveimg\\2",
1689                                 "\\1$secremoveimg\\2",
1690                                 "\\1$secremoveimg\\2",
1691                                 "\\1$secremoveimg\\2"
1692                                 )
1693                         ),
1694                 "/^href|action/i" =>
1695                     Array(
1696                           Array(
1697                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1698                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1699                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1700                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1701                                 ),
1702                           Array(
1703                                 "\\1#\\2",
1704                                 "\\1#\\2",
1705                                 "\\1#\\2",
1706                                 "\\1#\\2"
1707                                 )
1708                         ),
1709                 "/^style/i" =>
1710                     Array(
1711                           Array(
1712                                 "/expression/i",
1713                                 "/binding/i",
1714                                 "/behaviou*r/i",
1715                                 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1716                                 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1717                                 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1718                                 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1719                                ),
1720                           Array(
1721                                 "idiocy",
1722                                 "idiocy",
1723                                 "idiocy",
1724                                 "url(\\1#\\2)",
1725                                 "url(\\1#\\2)",
1726                                 "url(\\1#\\2)",
1727                                 "url(\\1#\\2)"
1728                                )
1729                           )
1730                 )
1731         );
1732     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1733         $view_unsafe_images = false;
1734     }
1735     if (!$view_unsafe_images){
1736         /**
1737          * Remove any references to http/https if view_unsafe_images set
1738          * to false.
1739          */
1740          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1741                     '/^([\'\"])\s*https*:.*([\'\"])/si');
1742          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1743                     "\\1$secremoveimg\\2");
1744          array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
1745                     '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1746          array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
1747                     "url(\\1$secremoveimg\\2)");
1748     }
1749
1750     $add_attr_to_tag = Array(
1751                              "/^a$/i" => Array('target'=>'"_new"')
1752                              );
1753     $trusted = sq_sanitize($body,
1754                            $tag_list,
1755                            $rm_tags_with_content,
1756                            $self_closing_tags,
1757                            $force_tag_closing,
1758                            $rm_attnames,
1759                            $bad_attvals,
1760                            $add_attr_to_tag,
1761                            $message,
1762                            $id,
1763                            $mailbox
1764                            );
1765     if (preg_match("|$secremoveimg|i", $trusted)){
1766         $has_unsafe_images = true;
1767     }
1768     return $trusted;
1769 }
1770
1771 ?>