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 $default_charset, $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     $encoded = false;
 564
 565     if (strtolower($default_charset) == 'iso-8859-1') {
 566         $string = str_replace("\240",' ',$string);
 567     }
 568
 569     $aString = explode(' ',$string);
 570     $ret = '';
 571     foreach ($aString as $chunk) {
 572         if ($encoded && !$chunk) {
 573             continue;
 574         } elseif (!$chunk) {
 575             $ret .= ' ';
 576             continue;
 577         }
 578         $encoded = false;
 579         /* if encoded words are not separated by a linear-space-white we still catch them */
 580         $j = $i-1;
 581         while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 582             /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
 583             if ($iLastMatch !== $j) {
 584                 if ($htmlsave) {
 585                     $ret .= '&nbsp;';
 586                 } else {
 587                     $ret .= ' ';
 588                 }
 589             }
 590             $iLastMatch = $i;
 591             $j = $i;
 592             $ret .= $res[1];
 593             $encoding = ucfirst($res[3]);
 594             switch ($encoding)
 595             {
 596             case 'B':
 597                 $replace = base64_decode($res[4]);
 598                 $ret .= charset_decode($res[2],$replace);
 599                 break;
 600             case 'Q':
 601                 $replace = str_replace('_', ' ', $res[4]);
 602                 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 603                                     $replace);
 604                 /* Only encode into entities by default. Some places
 605                  * don't need the encoding, like the compose form.
 606                  */
 607                 if ($utfencode) {
 608                     $replace = charset_decode($res[2], $replace);
 609                 } else {
 610                     if ($htmlsave) {
 611                         $replace = htmlspecialchars($replace);
 612                     }
 613                 }
 614                 $ret .= $replace;
 615                 break;
 616             default:
 617                 break;
 618             }
 619             $chunk = $res[5];
 620             $encoded = true;
 621         }
 622         if (!$encoded) {
 623             if ($htmlsave) {
 624                 $ret .= '&nbsp;';
 625             } else {
 626                 $ret .= ' ';
 627             }
 628         }
 629
 630         if (!$encoded && $htmlsave) {
 631             $ret .= htmlspecialchars($chunk);
 632         } else {
 633             $ret .= $chunk;
 634         }
 635         ++$i;
 636     }
 637     /* remove the first added space */
 638     if ($ret) {
 639         if ($htmlsave) {
 640             $ret = substr($ret,6);
 641         } else {
 642             $ret = substr($ret,1);
 643         }
 644     }
 645
 646     return $ret;
 647 }
 648
 649 /*
 650  * Encode a string according to RFC 1522 for use in headers if it
 651  * contains 8-bit characters or anything that looks like it should
 652  * be encoded.
 653  */
 654 function encodeHeader ($string) {
 655     global $default_charset, $languages, $squirrelmail_language;
 656
 657     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 658         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 659         return  $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
 660     }
 661     if (strtolower($default_charset) == 'iso-8859-1') {
 662         $string = str_replace("\240",' ',$string);
 663     }
 664
 665     // Encode only if the string contains 8-bit characters or =?
 666     $j = strlen($string);
 667     $max_l = 75 - strlen($default_charset) - 7;
 668     $aRet = array();
 669     $ret = '';
 670     $iEncStart = $enc_init = false;
 671     $cur_l = $iOffset = 0;
 672     for($i = 0; $i < $j; ++$i) {
 673         switch($string{$i})
 674         {
 675         case '=':
 676         case '<':
 677         case '>':
 678         case ',':
 679         case '?':
 680         case '_':
 681             if ($iEncStart === false) {
 682                 $iEncStart = $i;
 683             }
 684             $cur_l+=3;
 685             if ($cur_l > ($max_l-2)) {
 686                 /* if there is an stringpart that doesn't need encoding, add it */
 687                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 688                 $aRet[] = "=?$default_charset?Q?$ret?=";
 689                 $iOffset = $i;
 690                 $cur_l = 0;
 691                 $ret = '';
 692                 $iEncStart = false;
 693             } else {
 694                 $ret .= sprintf("=%02X",ord($string{$i}));
 695             }
 696             break;
 697         case '(':
 698         case ')':
 699             if ($iEncStart !== false) {
 700                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 701                 $aRet[] = "=?$default_charset?Q?$ret?=";
 702                 $iOffset = $i;
 703                 $cur_l = 0;
 704                 $ret = '';
 705                 $iEncStart = false;
 706             }
 707             break;
 708         case ' ':
 709             if ($iEncStart !== false) {
 710                 $cur_l++;
 711                 if ($cur_l > $max_l) {
 712                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 713                     $aRet[] = "=?$default_charset?Q?$ret?=";
 714                     $iOffset = $i;
 715                     $cur_l = 0;
 716                     $ret = '';
 717                     $iEncStart = false;
 718                 } else {
 719                     $ret .= '_';
 720                 }
 721             }
 722             break;
 723         default:
 724             $k = ord($string{$i});
 725             if ($k > 126) {
 726                 if ($iEncStart === false) {
 727                     $iEncStart = $i;
 728                 }
 729                 $cur_l += 3;
 730                 /* first we add the encoded string that reached it's max size */
 731                 if ($cur_l > ($max_l-2)) {
 732                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 733                     $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
 734                     $cur_l = 3;
 735                     $ret = '';
 736                     $iOffset = $i;
 737                     $iEncStart = $i;
 738                 }
 739                 $enc_init = true;
 740                 $ret .= sprintf("=%02X", $k);
 741             } else {
 742                 if ($iEncStart !== false) {
 743                     $cur_l++;
 744                     if ($cur_l > $max_l) {
 745                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 746                         $aRet[] = "=?$default_charset?Q?$ret?=";
 747                         $iEncStart = false;
 748                         $iOffset = $i;
 749                         $cur_l = 0;
 750                         $ret = '';
 751                     } else {
 752                         $ret .= $string{$i};
 753                     }
 754                 }
 755             }
 756             break;
 757         }
 758     }
 759
 760     if ($enc_init) {
 761         if ($iEncStart !== false) {
 762             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 763             $aRet[] = "=?$default_charset?Q?$ret?=";
 764         } else {
 765             $aRet[] = substr($string,$iOffset);
 766         }
 767         $string = implode('',$aRet);
 768     }
 769     return $string;
 770 }
 771
 772 /* This function trys to locate the entity_id of a specific mime element */
 773 function find_ent_id($id, $message) {
 774     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
 775         if ($message->entities[$i]->header->type0 == 'multipart')  {
 776             $ret = find_ent_id($id, $message->entities[$i]);
 777         } else {
 778             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
 779 //                if (sq_check_save_extension($message->entities[$i])) {
 780                     return $message->entities[$i]->entity_id;
 781 //                }
 782             }
 783         }
 784     }
 785     return $ret;
 786 }
 787
 788 function sq_check_save_extension($message) {
 789     $filename = $message->getFilename();
 790     $ext = substr($filename, strrpos($filename,'.')+1);
 791     $save_extensions = array('jpg','jpeg','gif','png','bmp');
 792     return in_array($ext, $save_extensions);
 793 }
 794
 795
 796 /**
 797  ** HTMLFILTER ROUTINES
 798  */
 799
 800 /**
 801  * This function returns the final tag out of the tag name, an array
 802  * of attributes, and the type of the tag. This function is called by
 803  * sq_sanitize internally.
 804  *
 805  * @param  $tagname  the name of the tag.
 806  * @param  $attary   the array of attributes and their values
 807  * @param  $tagtype  The type of the tag (see in comments).
 808  * @return           a string with the final tag representation.
 809  */
 810 function sq_tagprint($tagname, $attary, $tagtype){
 811     $me = 'sq_tagprint';
 812
 813     if ($tagtype == 2){
 814         $fulltag = '</' . $tagname . '>';
 815     } else {
 816         $fulltag = '<' . $tagname;
 817         if (is_array($attary) && sizeof($attary)){
 818             $atts = Array();
 819             while (list($attname, $attvalue) = each($attary)){
 820                 array_push($atts, "$attname=$attvalue");
 821             }
 822             $fulltag .= ' ' . join(" ", $atts);
 823         }
 824         if ($tagtype == 3){
 825             $fulltag .= ' /';
 826         }
 827         $fulltag .= '>';
 828     }
 829     return $fulltag;
 830 }
 831
 832 /**
 833  * A small helper function to use with array_walk. Modifies a by-ref
 834  * value and makes it lowercase.
 835  *
 836  * @param  $val a value passed by-ref.
 837  * @return      void since it modifies a by-ref value.
 838  */
 839 function sq_casenormalize(&$val){
 840     $val = strtolower($val);
 841 }
 842
 843 /**
 844  * This function skips any whitespace from the current position within
 845  * a string and to the next non-whitespace value.
 846  *
 847  * @param  $body   the string
 848  * @param  $offset the offset within the string where we should start
 849  *                 looking for the next non-whitespace character.
 850  * @return         the location within the $body where the next
 851  *                 non-whitespace char is located.
 852  */
 853 function sq_skipspace($body, $offset){
 854     $me = 'sq_skipspace';
 855     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
 856     if (sizeof($matches{1})){
 857         $count = strlen($matches{1});
 858         $offset += $count;
 859     }
 860     return $offset;
 861 }
 862
 863 /**
 864  * This function looks for the next character within a string.  It's
 865  * really just a glorified "strpos", except it catches if failures
 866  * nicely.
 867  *
 868  * @param  $body   The string to look for needle in.
 869  * @param  $offset Start looking from this position.
 870  * @param  $needle The character/string to look for.
 871  * @return         location of the next occurance of the needle, or
 872  *                 strlen($body) if needle wasn't found.
 873  */
 874 function sq_findnxstr($body, $offset, $needle){
 875     $me  = 'sq_findnxstr';
 876     $pos = strpos($body, $needle, $offset);
 877     if ($pos === FALSE){
 878         $pos = strlen($body);
 879     }
 880     return $pos;
 881 }
 882
 883 /**
 884  * This function takes a PCRE-style regexp and tries to match it
 885  * within the string.
 886  *
 887  * @param  $body   The string to look for needle in.
 888  * @param  $offset Start looking from here.
 889  * @param  $reg    A PCRE-style regex to match.
 890  * @return         Returns a false if no matches found, or an array
 891  *                 with the following members:
 892  *                 - integer with the location of the match within $body
 893  *                 - string with whatever content between offset and the match
 894  *                 - string with whatever it is we matched
 895  */
 896 function sq_findnxreg($body, $offset, $reg){
 897     $me = 'sq_findnxreg';
 898     $matches = Array();
 899     $retarr = Array();
 900     preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
 901     if (!$matches{0}){
 902         $retarr = false;
 903     } else {
 904         $retarr{0} = $offset + strlen($matches{1});
 905         $retarr{1} = $matches{1};
 906         $retarr{2} = $matches{2};
 907     }
 908     return $retarr;
 909 }
 910
 911 /**
 912  * This function looks for the next tag.
 913  *
 914  * @param  $body   String where to look for the next tag.
 915  * @param  $offset Start looking from here.
 916  * @return         false if no more tags exist in the body, or
 917  *                 an array with the following members:
 918  *                 - string with the name of the tag
 919  *                 - array with attributes and their values
 920  *                 - integer with tag type (1, 2, or 3)
 921  *                 - integer where the tag starts (starting "<")
 922  *                 - integer where the tag ends (ending ">")
 923  *                 first three members will be false, if the tag is invalid.
 924  */
 925 function sq_getnxtag($body, $offset){
 926     $me = 'sq_getnxtag';
 927     if ($offset > strlen($body)){
 928         return false;
 929     }
 930     $lt = sq_findnxstr($body, $offset, "<");
 931     if ($lt == strlen($body)){
 932         return false;
 933     }
 934     /**
 935      * We are here:
 936      * blah blah <tag attribute="value">
 937      * \---------^
 938      */
 939     $pos = sq_skipspace($body, $lt+1);
 940     if ($pos >= strlen($body)){
 941         return Array(false, false, false, $lt, strlen($body));
 942     }
 943     /**
 944      * There are 3 kinds of tags:
 945      * 1. Opening tag, e.g.:
 946      *    <a href="blah">
 947      * 2. Closing tag, e.g.:
 948      *    </a>
 949      * 3. XHTML-style content-less tag, e.g.:
 950      *    <img src="blah"/>
 951      */
 952     $tagtype = false;
 953     switch (substr($body, $pos, 1)){
 954         case '/':
 955             $tagtype = 2;
 956             $pos++;
 957             break;
 958         case '!':
 959             /**
 960              * A comment or an SGML declaration.
 961              */
 962             if (substr($body, $pos+1, 2) == "--"){
 963                 $gt = strpos($body, "-->", $pos);
 964                 if ($gt === false){
 965                     $gt = strlen($body);
 966                 } else {
 967                     $gt += 2;
 968                 }
 969                 return Array(false, false, false, $lt, $gt);
 970             } else {
 971                 $gt = sq_findnxstr($body, $pos, ">");
 972                 return Array(false, false, false, $lt, $gt);
 973             }
 974             break;
 975         default:
 976             /**
 977              * Assume tagtype 1 for now. If it's type 3, we'll switch values
 978              * later.
 979              */
 980             $tagtype = 1;
 981             break;
 982     }
 983
 984     $tag_start = $pos;
 985     $tagname = '';
 986     /**
 987      * Look for next [\W-_], which will indicate the end of the tag name.
 988      */
 989     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
 990     if ($regary == false){
 991         return Array(false, false, false, $lt, strlen($body));
 992     }
 993     list($pos, $tagname, $match) = $regary;
 994     $tagname = strtolower($tagname);
 995
 996     /**
 997      * $match can be either of these:
 998      * '>'  indicating the end of the tag entirely.
 999      * '\s' indicating the end of the tag name.
1000      * '/'  indicating that this is type-3 xhtml tag.
1001      *
1002      * Whatever else we find there indicates an invalid tag.
1003      */
1004     switch ($match){
1005         case '/':
1006             /**
1007              * This is an xhtml-style tag with a closing / at the
1008              * end, like so: <img src="blah"/>. Check if it's followed
1009              * by the closing bracket. If not, then this tag is invalid
1010              */
1011             if (substr($body, $pos, 2) == "/>"){
1012                 $pos++;
1013                 $tagtype = 3;
1014             } else {
1015                 $gt = sq_findnxstr($body, $pos, ">");
1016                 $retary = Array(false, false, false, $lt, $gt);
1017                 return $retary;
1018             }
1019         case '>':
1020             return Array($tagname, false, $tagtype, $lt, $pos);
1021             break;
1022         default:
1023             /**
1024              * Check if it's whitespace
1025              */
1026             if (!preg_match('/\s/', $match)){
1027                 /**
1028                  * This is an invalid tag! Look for the next closing ">".
1029                  */
1030                 $gt = sq_findnxstr($body, $offset, ">");
1031                 return Array(false, false, false, $lt, $gt);
1032             }
1033             break;
1034     }
1035
1036     /**
1037      * At this point we're here:
1038      * <tagname  attribute='blah'>
1039      * \-------^
1040      *
1041      * At this point we loop in order to find all attributes.
1042      */
1043     $attname = '';
1044     $atttype = false;
1045     $attary = Array();
1046
1047     while ($pos <= strlen($body)){
1048         $pos = sq_skipspace($body, $pos);
1049         if ($pos == strlen($body)){
1050             /**
1051              * Non-closed tag.
1052              */
1053             return Array(false, false, false, $lt, $pos);
1054         }
1055         /**
1056          * See if we arrived at a ">" or "/>", which means that we reached
1057          * the end of the tag.
1058          */
1059         $matches = Array();
1060         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1061             /**
1062              * Yep. So we did.
1063              */
1064             $pos += strlen($matches{1});
1065             if ($matches{2} == "/>"){
1066                 $tagtype = 3;
1067                 $pos++;
1068             }
1069             return Array($tagname, $attary, $tagtype, $lt, $pos);
1070         }
1071
1072         /**
1073          * There are several types of attributes, with optional
1074          * [:space:] between members.
1075          * Type 1:
1076          *   attrname[:space:]=[:space:]'CDATA'
1077          * Type 2:
1078          *   attrname[:space:]=[:space:]"CDATA"
1079          * Type 3:
1080          *   attr[:space:]=[:space:]CDATA
1081          * Type 4:
1082          *   attrname
1083          *
1084          * We leave types 1 and 2 the same, type 3 we check for
1085          * '"' and convert to "&quot" if needed, then wrap in
1086          * double quotes. Type 4 we convert into:
1087          * attrname="yes".
1088          */
1089         $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1090         if ($regary == false){
1091             /**
1092              * Looks like body ended before the end of tag.
1093              */
1094             return Array(false, false, false, $lt, strlen($body));
1095         }
1096         list($pos, $attname, $match) = $regary;
1097         $attname = strtolower($attname);
1098         /**
1099          * We arrived at the end of attribute name. Several things possible
1100          * here:
1101          * '>'  means the end of the tag and this is attribute type 4
1102          * '/'  if followed by '>' means the same thing as above
1103          * '\s' means a lot of things -- look what it's followed by.
1104          *      anything else means the attribute is invalid.
1105          */
1106         switch($match){
1107             case '/':
1108                 /**
1109                  * This is an xhtml-style tag with a closing / at the
1110                  * end, like so: <img src="blah"/>. Check if it's followed
1111                  * by the closing bracket. If not, then this tag is invalid
1112                  */
1113                 if (substr($body, $pos, 2) == "/>"){
1114                     $pos++;
1115                     $tagtype = 3;
1116                 } else {
1117                     $gt = sq_findnxstr($body, $pos, ">");
1118                     $retary = Array(false, false, false, $lt, $gt);
1119                     return $retary;
1120                 }
1121             case '>':
1122                 $attary{$attname} = '"yes"';
1123                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1124                 break;
1125             default:
1126                 /**
1127                  * Skip whitespace and see what we arrive at.
1128                  */
1129                 $pos = sq_skipspace($body, $pos);
1130                 $char = substr($body, $pos, 1);
1131                 /**
1132                  * Two things are valid here:
1133                  * '=' means this is attribute type 1 2 or 3.
1134                  * \w means this was attribute type 4.
1135                  * anything else we ignore and re-loop. End of tag and
1136                  * invalid stuff will be caught by our checks at the beginning
1137                  * of the loop.
1138                  */
1139                 if ($char == "="){
1140                     $pos++;
1141                     $pos = sq_skipspace($body, $pos);
1142                     /**
1143                      * Here are 3 possibilities:
1144                      * "'"  attribute type 1
1145                      * '"'  attribute type 2
1146                      * everything else is the content of tag type 3
1147                      */
1148                     $quot = substr($body, $pos, 1);
1149                     if ($quot == "'"){
1150                         $regary = sq_findnxreg($body, $pos+1, "\'");
1151                         if ($regary == false){
1152                             return Array(false, false, false, $lt, strlen($body));
1153                         }
1154                         list($pos, $attval, $match) = $regary;
1155                         $pos++;
1156                         $attary{$attname} = "'" . $attval . "'";
1157                     } else if ($quot == '"'){
1158                         $regary = sq_findnxreg($body, $pos+1, '\"');
1159                         if ($regary == false){
1160                             return Array(false, false, false, $lt, strlen($body));
1161                         }
1162                         list($pos, $attval, $match) = $regary;
1163                         $pos++;
1164                         $attary{$attname} = '"' . $attval . '"';
1165                     } else {
1166                         /**
1167                          * These are hateful. Look for \s, or >.
1168                          */
1169                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1170                         if ($regary == false){
1171                             return Array(false, false, false, $lt, strlen($body));
1172                         }
1173                         list($pos, $attval, $match) = $regary;
1174                         /**
1175                          * If it's ">" it will be caught at the top.
1176                          */
1177                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1178                         $attary{$attname} = '"' . $attval . '"';
1179                     }
1180                 } else if (preg_match("|[\w/>]|", $char)) {
1181                     /**
1182                      * That was attribute type 4.
1183                      */
1184                     $attary{$attname} = '"yes"';
1185                 } else {
1186                     /**
1187                      * An illegal character. Find next '>' and return.
1188                      */
1189                     $gt = sq_findnxstr($body, $pos, ">");
1190                     return Array(false, false, false, $lt, $gt);
1191                 }
1192                 break;
1193         }
1194     }
1195     /**
1196      * The fact that we got here indicates that the tag end was never
1197      * found. Return invalid tag indication so it gets stripped.
1198      */
1199     return Array(false, false, false, $lt, strlen($body));
1200 }
1201
1202 /**
1203  * This function checks attribute values for entity-encoded values
1204  * and returns them translated into 8-bit strings so we can run
1205  * checks on them.
1206  *
1207  * @param  $attvalue A string to run entity check against.
1208  * @return           Translated value.
1209  */
1210 function sq_deent($attvalue){
1211     $me = 'sq_deent';
1212     /**
1213      * See if we have to run the checks first. All entities must start
1214      * with "&".
1215      */
1216     if (strpos($attvalue, "&") === false){
1217         return $attvalue;
1218     }
1219     /**
1220      * Check named entities first.
1221      */
1222     $trans = get_html_translation_table(HTML_ENTITIES);
1223     /**
1224      * Leave &quot; in, as it can mess us up.
1225      */
1226     $trans = array_flip($trans);
1227     unset($trans{"&quot;"});
1228     while (list($ent, $val) = each($trans)){
1229         $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1230     }
1231     /**
1232      * Now translate numbered entities from 1 to 255 if needed.
1233      */
1234     if (strpos($attvalue, "#") !== false){
1235         $omit = Array(34, 39);
1236         for ($asc=1; $asc<256; $asc++){
1237             if (!in_array($asc, $omit)){
1238                 $chr = chr($asc);
1239                 $attvalue = preg_replace("/\&#0*$asc;*(\D)/si", "$chr\\1",
1240                                          $attvalue);
1241                 $attvalue = preg_replace("/\&#x0*".dechex($asc).";*(\W)/si",
1242                                          "$chr\\1", $attvalue);
1243             }
1244         }
1245     }
1246     return $attvalue;
1247 }
1248
1249 /**
1250  * This function runs various checks against the attributes.
1251  *
1252  * @param  $tagname         String with the name of the tag.
1253  * @param  $attary          Array with all tag attributes.
1254  * @param  $rm_attnames     See description for sq_sanitize
1255  * @param  $bad_attvals     See description for sq_sanitize
1256  * @param  $add_attr_to_tag See description for sq_sanitize
1257  * @param  $message         message object
1258  * @param  $id              message id
1259  * @return                  Array with modified attributes.
1260  */
1261 function sq_fixatts($tagname,
1262                     $attary,
1263                     $rm_attnames,
1264                     $bad_attvals,
1265                     $add_attr_to_tag,
1266                     $message,
1267                     $id,
1268                     $mailbox
1269                     ){
1270     $me = 'sq_fixatts';
1271     while (list($attname, $attvalue) = each($attary)){
1272         /**
1273          * See if this attribute should be removed.
1274          */
1275         foreach ($rm_attnames as $matchtag=>$matchattrs){
1276             if (preg_match($matchtag, $tagname)){
1277                 foreach ($matchattrs as $matchattr){
1278                     if (preg_match($matchattr, $attname)){
1279                         unset($attary{$attname});
1280                         continue;
1281                     }
1282                 }
1283             }
1284         }
1285         /**
1286          * Remove any entities.
1287          */
1288         $attvalue = sq_deent($attvalue);
1289
1290         /**
1291          * Now let's run checks on the attvalues.
1292          * I don't expect anyone to comprehend this. If you do,
1293          * get in touch with me so I can drive to where you live and
1294          * shake your hand personally. :)
1295          */
1296         foreach ($bad_attvals as $matchtag=>$matchattrs){
1297             if (preg_match($matchtag, $tagname)){
1298                 foreach ($matchattrs as $matchattr=>$valary){
1299                     if (preg_match($matchattr, $attname)){
1300                         /**
1301                          * There are two arrays in valary.
1302                          * First is matches.
1303                          * Second one is replacements
1304                          */
1305                         list($valmatch, $valrepl) = $valary;
1306                         $newvalue =
1307                             preg_replace($valmatch, $valrepl, $attvalue);
1308                         if ($newvalue != $attvalue){
1309                             $attary{$attname} = $newvalue;
1310                         }
1311                     }
1312                 }
1313             }
1314         }
1315         /**
1316          * Turn cid: urls into http-friendly ones.
1317          */
1318         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1319             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1320         }
1321     }
1322     /**
1323      * See if we need to append any attributes to this tag.
1324      */
1325     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1326         if (preg_match($matchtag, $tagname)){
1327             $attary = array_merge($attary, $addattary);
1328         }
1329     }
1330     return $attary;
1331 }
1332
1333 /**
1334  * This function edits the style definition to make them friendly and
1335  * usable in squirrelmail.
1336  *
1337  * @param  $message  the message object
1338  * @param  $id       the message id
1339  * @param  $content  a string with whatever is between <style> and </style>
1340  * @return           a string with edited content.
1341  */
1342 function sq_fixstyle($message, $id, $content){
1343     global $view_unsafe_images;
1344     $me = 'sq_fixstyle';
1345     /**
1346      * First look for general BODY style declaration, which would be
1347      * like so:
1348      * body {background: blah-blah}
1349      * and change it to .bodyclass so we can just assign it to a <div>
1350      */
1351     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1352     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1353     /**
1354      * Fix url('blah') declarations.
1355      */
1356     $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1357                             "url(\\1$secremoveimg\\2)", $content);
1358     /**
1359      * Fix url('https*://.*) declarations but only if $view_unsafe_images
1360      * is false.
1361      */
1362     if (!$view_unsafe_images){
1363         $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1364                                 "url(\\1$secremoveimg\\2)", $content);
1365     }
1366
1367     /**
1368      * Fix urls that refer to cid:
1369      */
1370     while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1371                       $matches)){
1372         $cidurl = $matches{1};
1373         $httpurl = sq_cid2http($message, $id, $cidurl);
1374         $content = preg_replace("|url\($cidurl\)|si",
1375                                 "url($httpurl)", $content);
1376     }
1377
1378     /**
1379      * Fix stupid css declarations which lead to vulnerabilities
1380      * in IE.
1381      */
1382     $match   = Array('/expression/i',
1383                      '/behaviou*r/i',
1384                      '/binding/i');
1385     $replace = Array('idiocy', 'idiocy', 'idiocy');
1386     $content = preg_replace($match, $replace, $content);
1387     return $content;
1388 }
1389
1390 /**
1391  * This function converts cid: url's into the ones that can be viewed in
1392  * the browser.
1393  *
1394  * @param  $message  the message object
1395  * @param  $id       the message id
1396  * @param  $cidurl   the cid: url.
1397  * @return           a string with a http-friendly url
1398  */
1399 function sq_cid2http($message, $id, $cidurl, $mailbox){
1400     /**
1401      * Get rid of quotes.
1402      */
1403     $quotchar = substr($cidurl, 0, 1);
1404     $cidurl = str_replace($quotchar, "", $cidurl);
1405     $cidurl = substr(trim($cidurl), 4);
1406     $linkurl = find_ent_id($cidurl, $message);
1407     /* in case of non-save cid links $httpurl should be replaced by a sort of
1408        unsave link image */
1409     $httpurl = '';
1410     if ($linkurl) {
1411         $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1412                    "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1413                    '&amp;ent_id=' . $linkurl . $quotchar;
1414     }
1415     return $httpurl;
1416 }
1417
1418 /**
1419  * This function changes the <body> tag into a <div> tag since we
1420  * can't really have a body-within-body.
1421  *
1422  * @param  $attary  an array of attributes and values of <body>
1423  * @return          a modified array of attributes to be set for <div>
1424  */
1425 function sq_body2div($attary){
1426     $me = 'sq_body2div';
1427     $divattary = Array('class' => "'bodyclass'");
1428     $bgcolor = '#ffffff';
1429     $text = '#000000';
1430     $styledef = '';
1431     if (is_array($attary) && sizeof($attary) > 0){
1432         foreach ($attary as $attname=>$attvalue){
1433             $quotchar = substr($attvalue, 0, 1);
1434             $attvalue = str_replace($quotchar, "", $attvalue);
1435             switch ($attname){
1436                 case 'background':
1437                     $styledef .= "background-image: url('$attvalue'); ";
1438                     break;
1439                 case 'bgcolor':
1440                     $styledef .= "background-color: $attvalue; ";
1441                     break;
1442                 case 'text':
1443                     $styledef .= "color: $attvalue; ";
1444                     break;
1445             }
1446         }
1447         if (strlen($styledef) > 0){
1448             $divattary{"style"} = "\"$styledef\"";
1449         }
1450     }
1451     return $divattary;
1452 }
1453
1454 /**
1455  * This is the main function and the one you should actually be calling.
1456  * There are several variables you should be aware of an which need
1457  * special description.
1458  *
1459  * Since the description is quite lengthy, see it here:
1460  * http://www.mricon.com/html/phpfilter.html
1461  *
1462  * @param $body                 the string with HTML you wish to filter
1463  * @param $tag_list             see description above
1464  * @param $rm_tags_with_content see description above
1465  * @param $self_closing_tags    see description above
1466  * @param $force_tag_closing    see description above
1467  * @param $rm_attnames          see description above
1468  * @param $bad_attvals          see description above
1469  * @param $add_attr_to_tag      see description above
1470  * @param $message              message object
1471  * @param $id                   message id
1472  * @return                      sanitized html safe to show on your pages.
1473  */
1474 function sq_sanitize($body,
1475                      $tag_list,
1476                      $rm_tags_with_content,
1477                      $self_closing_tags,
1478                      $force_tag_closing,
1479                      $rm_attnames,
1480                      $bad_attvals,
1481                      $add_attr_to_tag,
1482                      $message,
1483                      $id,
1484                      $mailbox
1485                      ){
1486     $me = 'sq_sanitize';
1487     /**
1488      * Normalize rm_tags and rm_tags_with_content.
1489      */
1490     @array_walk($rm_tags, 'sq_casenormalize');
1491     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1492     @array_walk($self_closing_tags, 'sq_casenormalize');
1493     /**
1494      * See if tag_list is of tags to remove or tags to allow.
1495      * false  means remove these tags
1496      * true   means allow these tags
1497      */
1498     $rm_tags = array_shift($tag_list);
1499     $curpos = 0;
1500     $open_tags = Array();
1501     $trusted = "<!-- begin sanitized html -->\n";
1502     $skip_content = false;
1503     /**
1504      * Take care of netscape's stupid javascript entities like
1505      * &{alert('boo')};
1506      */
1507     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1508
1509     while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1510         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1511         $free_content = substr($body, $curpos, $lt-$curpos);
1512         /**
1513          * Take care of <style>
1514          */
1515         if ($tagname == "style" && $tagtype == 2){
1516             /**
1517              * This is a closing </style>. Edit the
1518              * content before we apply it.
1519              */
1520             $free_content = sq_fixstyle($message, $id, $free_content);
1521         }
1522         if ($skip_content == false){
1523             $trusted .= $free_content;
1524         }
1525         if ($tagname != FALSE){
1526             if ($tagtype == 2){
1527                 if ($skip_content == $tagname){
1528                     /**
1529                      * Got to the end of tag we needed to remove.
1530                      */
1531                     $tagname = false;
1532                     $skip_content = false;
1533                 } else {
1534                     if ($skip_content == false){
1535                         if ($tagname == "body"){
1536                             $tagname = "div";
1537                         } else {
1538                             if (isset($open_tags{$tagname}) &&
1539                                 $open_tags{$tagname} > 0){
1540                                 $open_tags{$tagname}--;
1541                             } else {
1542                                 $tagname = false;
1543                             }
1544                         }
1545                     }
1546                 }
1547             } else {
1548                 /**
1549                  * $rm_tags_with_content
1550                  */
1551                 if ($skip_content == false){
1552                     /**
1553                      * See if this is a self-closing type and change
1554                      * tagtype appropriately.
1555                      */
1556                     if ($tagtype == 1
1557                         && in_array($tagname, $self_closing_tags)){
1558                         $tagtype=3;
1559                     }
1560                     /**
1561                      * See if we should skip this tag and any content
1562                      * inside it.
1563                      */
1564                     if ($tagtype == 1 &&
1565                         in_array($tagname, $rm_tags_with_content)){
1566                         $skip_content = $tagname;
1567                     } else {
1568                         if (($rm_tags == false
1569                              && in_array($tagname, $tag_list)) ||
1570                             ($rm_tags == true &&
1571                              !in_array($tagname, $tag_list))){
1572                             $tagname = false;
1573                         } else {
1574                             if ($tagtype == 1){
1575                                 if (isset($open_tags{$tagname})){
1576                                     $open_tags{$tagname}++;
1577                                 } else {
1578                                     $open_tags{$tagname}=1;
1579                                 }
1580                             }
1581                             /**
1582                              * This is where we run other checks.
1583                              */
1584                             if (is_array($attary) && sizeof($attary) > 0){
1585                                 $attary = sq_fixatts($tagname,
1586                                                      $attary,
1587                                                      $rm_attnames,
1588                                                      $bad_attvals,
1589                                                      $add_attr_to_tag,
1590                                                      $message,
1591                                                      $id,
1592                                                      $mailbox
1593                                                      );
1594                             }
1595                             /**
1596                              * Convert body into div.
1597                              */
1598                             if ($tagname == "body"){
1599                                 $tagname = "div";
1600                                 $attary = sq_body2div($attary, $message, $id);
1601                             }
1602                         }
1603                     }
1604                 }
1605             }
1606             if ($tagname != false && $skip_content == false){
1607                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1608             }
1609         }
1610         $curpos = $gt+1;
1611     }
1612     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1613     if ($force_tag_closing == true){
1614         foreach ($open_tags as $tagname=>$opentimes){
1615             while ($opentimes > 0){
1616                 $trusted .= '</' . $tagname . '>';
1617                 $opentimes--;
1618             }
1619         }
1620         $trusted .= "\n";
1621     }
1622     $trusted .= "<!-- end sanitized html -->\n";
1623     return $trusted;
1624 }
1625
1626 /**
1627  * This is a wrapper function to call html sanitizing routines.
1628  *
1629  * @param  $body  the body of the message
1630  * @param  $id    the id of the message
1631  * @return        a string with html safe to display in the browser.
1632  */
1633 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1634     global $attachment_common_show_images, $view_unsafe_images,
1635            $has_unsafe_images;
1636     /**
1637      * Don't display attached images in HTML mode.
1638      */
1639     $attachment_common_show_images = false;
1640     $tag_list = Array(
1641                       false,
1642                       "object",
1643                       "meta",
1644                       "html",
1645                       "head",
1646                       "base",
1647                       "link",
1648                       "frame",
1649                       "iframe"
1650                       );
1651
1652     $rm_tags_with_content = Array(
1653                                   "script",
1654                                   "applet",
1655                                   "embed",
1656                                   "title"
1657                                   );
1658
1659     $self_closing_tags =  Array(
1660                                 "img",
1661                                 "br",
1662                                 "hr",
1663                                 "input"
1664                                 );
1665
1666     $force_tag_closing = false;
1667
1668     $rm_attnames = Array(
1669                          "/.*/" =>
1670                          Array(
1671                                "/target/i",
1672                                "/^on.*/i",
1673                                "/^dynsrc/i",
1674                                "/^data.*/i",
1675                                "/^lowsrc.*/i"
1676                                )
1677                          );
1678
1679     $secremoveimg = "../images/" . _("sec_remove_eng.png");
1680     $bad_attvals = Array(
1681         "/.*/" =>
1682             Array(
1683                 "/^src|background/i" =>
1684                     Array(
1685                           Array(
1686                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1687                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1688                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1689                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1690                                 ),
1691                           Array(
1692                                 "\\1$secremoveimg\\2",
1693                                 "\\1$secremoveimg\\2",
1694                                 "\\1$secremoveimg\\2",
1695                                 "\\1$secremoveimg\\2"
1696                                 )
1697                         ),
1698                 "/^href|action/i" =>
1699                     Array(
1700                           Array(
1701                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1702                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1703                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1704                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1705                                 ),
1706                           Array(
1707                                 "\\1#\\2",
1708                                 "\\1#\\2",
1709                                 "\\1#\\2",
1710                                 "\\1#\\2"
1711                                 )
1712                         ),
1713                 "/^style/i" =>
1714                     Array(
1715                           Array(
1716                                 "/expression/i",
1717                                 "/binding/i",
1718                                 "/behaviou*r/i",
1719                                 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1720                                 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1721                                 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1722                                 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1723                                ),
1724                           Array(
1725                                 "idiocy",
1726                                 "idiocy",
1727                                 "idiocy",
1728                                 "url(\\1#\\2)",
1729                                 "url(\\1#\\2)",
1730                                 "url(\\1#\\2)",
1731                                 "url(\\1#\\2)"
1732                                )
1733                           )
1734                 )
1735         );
1736     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1737         $view_unsafe_images = false;
1738     }
1739     if (!$view_unsafe_images){
1740         /**
1741          * Remove any references to http/https if view_unsafe_images set
1742          * to false.
1743          */
1744          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1745                     '/^([\'\"])\s*https*:.*([\'\"])/si');
1746          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1747                     "\\1$secremoveimg\\2");
1748          array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
1749                     '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1750          array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
1751                     "url(\\1$secremoveimg\\2)");
1752     }
1753
1754     $add_attr_to_tag = Array(
1755                              "/^a$/i" => Array('target'=>'"_new"')
1756                              );
1757     $trusted = sq_sanitize($body,
1758                            $tag_list,
1759                            $rm_tags_with_content,
1760                            $self_closing_tags,
1761                            $force_tag_closing,
1762                            $rm_attnames,
1763                            $bad_attvals,
1764                            $add_attr_to_tag,
1765                            $message,
1766                            $id,
1767                            $mailbox
1768                            );
1769     if (preg_match("|$secremoveimg|i", $trusted)){
1770         $has_unsafe_images = true;
1771     }
1772     return $trusted;
1773 }
1774
1775 ?>