functions/mime.php

   1 <?php
   2
   3 /**
   4  * mime.php
   5  *
   6  * Copyright (c) 1999-2003 The SquirrelMail Project Team
   7  * Licensed under the GNU GPL. For full terms see the file COPYING.
   8  *
   9  * This contains the functions necessary to detect and decode MIME
  10  * messages.
  11  *
  12  * $Id$
  13  */
  14
  15 require_once(SM_PATH . 'functions/imap.php');
  16 require_once(SM_PATH . 'functions/attachment_common.php');
  17
  18 /* --------------------------------------------------------------------------------- */
  19 /* MIME DECODING                                                                     */
  20 /* --------------------------------------------------------------------------------- */
  21
  22 /* This function gets the structure of a message and stores it in the "message" class.
  23  * It will return this object for use with all relevant header information and
  24  * fully parsed into the standard "message" object format.
  25  */
  26
  27 function mime_structure ($bodystructure, $flags=array()) {
  28
  29     /* Isolate the body structure and remove beginning and end parenthesis. */
  30     $read = trim(substr ($bodystructure, strpos(strtolower($bodystructure), 'bodystructure') + 13));
  31     $read = trim(substr ($read, 0, -1));
  32     $i = 0;
  33     $msg = Message::parseStructure($read,$i);
  34     if (!is_object($msg)) {
  35         include_once(SM_PATH . 'functions/display_messages.php');
  36         global $color, $mailbox;
  37         displayPageHeader( $color, urldecode($mailbox) );
  38         echo "<BODY TEXT=\"$color[8]\" BGCOLOR=\"$color[4]\" LINK=\"$color[7]\" VLINK=\"$color[7]\" ALINK=\"$color[7]\">\n\n" .
  39          '<CENTER>';
  40         $errormessage  = _("SquirrelMail could not decode the bodystructure of the message");
  41         $errormessage .= '<BR>'._("the provided bodystructure by your imap-server").':<BR><BR>';
  42         $errormessage .= '<table><tr><td>' . htmlspecialchars($read) . '</td></tr></table>';
  43         plain_error_message( $errormessage, $color );
  44         echo '</body></html>';
  45         exit;
  46     }
  47     if (count($flags)) {
  48         foreach ($flags as $flag) {
  49             $char = strtoupper($flag{1});
  50             switch ($char) {
  51                 case 'S':
  52                     if (strtolower($flag) == '\\seen') {
  53                         $msg->is_seen = true;
  54                     }
  55                     break;
  56                 case 'A':
  57                     if (strtolower($flag) == '\\answered') {
  58                         $msg->is_answered = true;
  59                     }
  60                     break;
  61                 case 'D':
  62                     if (strtolower($flag) == '\\deleted') {
  63                         $msg->is_deleted = true;
  64                     }
  65                     break;
  66                 case 'F':
  67                     if (strtolower($flag) == '\\flagged') {
  68                         $msg->is_flagged = true;
  69                     }
  70                     break;
  71                 case 'M':
  72                     if (strtolower($flag) == '$mdnsent') {
  73                         $msg->is_mdnsent = true;
  74                     }
  75                     break;
  76                 default:
  77                     break;
  78             }
  79         }
  80     }
  81     //    listEntities($msg);
  82     return $msg;
  83 }
  84
  85
  86
  87 /* This starts the parsing of a particular structure.  It is called recursively,
  88  * so it can be passed different structures.  It returns an object of type
  89  * $message.
  90  * First, it checks to see if it is a multipart message.  If it is, then it
  91  * handles that as it sees is necessary.  If it is just a regular entity,
  92  * then it parses it and adds the necessary header information (by calling out
  93  * to mime_get_elements()
  94  */
  95
  96 function mime_fetch_body($imap_stream, $id, $ent_id=1) {
  97     global $uid_support;
  98     /* Do a bit of error correction.  If we couldn't find the entity id, just guess
  99      * that it is the first one.  That is usually the case anyway.
 100      */
 101     if (!$ent_id) {
 102         $cmd = "FETCH $id BODY[]";
 103     } else {
 104         $cmd = "FETCH $id BODY[$ent_id]";
 105     }
 106
 107     $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, $uid_support);
 108     do {
 109         $topline = trim(array_shift($data));
 110     } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
 111
 112     $wholemessage = implode('', $data);
 113     if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
 114         $ret = substr($wholemessage, 0, $regs[1]);
 115         /* There is some information in the content info header that could be important
 116          * in order to parse html messages. Let's get them here.
 117          */
 118         if ($ret{0} == '<') {
 119             $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, $uid_support);
 120         }
 121     } else if (ereg('"([^"]*)"', $topline, $regs)) {
 122         $ret = $regs[1];
 123     } else {
 124         global $where, $what, $mailbox, $passed_id, $startMessage;
 125         $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
 126         if (isset($where) && isset($what)) {
 127             $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
 128         } else {
 129             $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 130         }
 131         $par .= '&amp;response=' . urlencode($response) .
 132                 '&amp;message='  . urlencode($message)  .
 133                 '&amp;topline='  . urlencode($topline);
 134
 135         echo   '<tt><br>' .
 136                '<table width="80%"><tr>' .
 137                '<tr><td colspan=2>' .
 138                _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
 139                '</td></tr>' .
 140                '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
 141                '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
 142                '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
 143                '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
 144                "</table><BR></tt></font><hr>";
 145
 146         $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, $uid_support);
 147         array_shift($data);
 148         $wholemessage = implode('', $data);
 149
 150         $ret = $wholemessage;
 151     }
 152     return $ret;
 153 }
 154
 155 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding) {
 156     global $uid_support;
 157
 158     $sid = sqimap_session_id($uid_support);
 159     /* Don't kill the connection if the browser is over a dialup
 160      * and it would take over 30 seconds to download it.
 161      * Don´t call set_time_limit in safe mode.
 162      */
 163
 164     if (!ini_get('safe_mode')) {
 165         set_time_limit(0);
 166     }
 167     if ($uid_support) {
 168        $sid_s = substr($sid,0,strpos($sid, ' '));
 169     } else {
 170        $sid_s = $sid;
 171     }
 172
 173     $body = mime_fetch_body ($imap_stream, $id, $ent_id);
 174     echo decodeBody($body, $encoding);
 175     return;
 176 /*
 177     fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
 178     $cnt = 0;
 179     $continue = true;
 180     $read = fgets ($imap_stream,8192);
 181
 182
 183     // This could be bad -- if the section has sqimap_session_id() . ' OK'
 184     // or similar, it will kill the download.
 185     while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
 186         if (trim($read) == ')==') {
 187             $read1 = $read;
 188             $read = fgets ($imap_stream,4096);
 189             if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
 190                 return;
 191             } else {
 192                 echo decodeBody($read1, $encoding) .
 193                      decodeBody($read, $encoding);
 194             }
 195         } else if ($cnt) {
 196             echo decodeBody($read, $encoding);
 197         }
 198         $read = fgets ($imap_stream,4096);
 199         $cnt++;
 200 //      break;
 201     }
 202 */
 203 }
 204
 205 /* -[ END MIME DECODING ]----------------------------------------------------------- */
 206
 207 /* This is here for debugging purposes.  It will print out a list
 208  * of all the entity IDs that are in the $message object.
 209  */
 210 function listEntities ($message) {
 211     if ($message) {
 212         echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br>';
 213         for ($i = 0; isset($message->entities[$i]); $i++) {
 214             echo "$i : ";
 215             $msg = listEntities($message->entities[$i]);
 216
 217             if ($msg) {
 218                 echo "return: ";
 219                 return $msg;
 220             }
 221         }
 222     }
 223 }
 224
 225 function getPriorityStr($priority) {
 226     $priority_level = substr($priority,0,1);
 227
 228     switch($priority_level) {
 229         /* Check for a higher then normal priority. */
 230         case '1':
 231         case '2':
 232             $priority_string = _("High");
 233             break;
 234
 235         /* Check for a lower then normal priority. */
 236         case '4':
 237         case '5':
 238             $priority_string = _("Low");
 239             break;
 240
 241         /* Check for a normal priority. */
 242         case '3':
 243         default:
 244             $priority_level = '3';
 245             $priority_string = _("Normal");
 246             break;
 247
 248     }
 249     return $priority_string;
 250 }
 251
 252 /* returns a $message object for a particular entity id */
 253 function getEntity ($message, $ent_id) {
 254     return $message->getEntity($ent_id);
 255 }
 256
 257 /* translateText
 258  * Extracted from strings.php 23/03/2002
 259  */
 260
 261 function translateText(&$body, $wrap_at, $charset) {
 262     global $where, $what;   /* from searching */
 263     global $color;          /* color theme */
 264
 265     require_once(SM_PATH . 'functions/url_parser.php');
 266
 267     $body_ary = explode("\n", $body);
 268     for ($i=0; $i < count($body_ary); $i++) {
 269         $line = $body_ary[$i];
 270         if (strlen($line) - 2 >= $wrap_at) {
 271             sqWordWrap($line, $wrap_at);
 272         }
 273         $line = charset_decode($charset, $line);
 274         $line = str_replace("\t", '        ', $line);
 275
 276         parseUrl ($line);
 277
 278         $quotes = 0;
 279         $pos = 0;
 280         $j = strlen($line);
 281
 282         while ($pos < $j) {
 283             if ($line[$pos] == ' ') {
 284                 $pos++;
 285             } else if (strpos($line, '&gt;', $pos) === $pos) {
 286                 $pos += 4;
 287                 $quotes++;
 288             } else {
 289                 break;
 290             }
 291         }
 292
 293         if ($quotes % 2) {
 294             if (!isset($color[13])) {
 295                 $color[13] = '#800000';
 296             }
 297             $line = '<font color="' . $color[13] . '">' . $line . '</font>';
 298         } elseif ($quotes) {
 299             if (!isset($color[14])) {
 300                 $color[14] = '#FF0000';
 301             }
 302             $line = '<font color="' . $color[14] . '">' . $line . '</font>';
 303         }
 304
 305         $body_ary[$i] = $line;
 306     }
 307     $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
 308 }
 309
 310 /* This returns a parsed string called $body. That string can then
 311  * be displayed as the actual message in the HTML. It contains
 312  * everything needed, including HTML Tags, Attachments at the
 313  * bottom, etc.
 314  */
 315 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
 316     /* This if statement checks for the entity to show as the
 317      * primary message. To add more of them, just put them in the
 318      * order that is their priority.
 319      */
 320     global $startMessage, $username, $key, $imapServerAddress, $imapPort,
 321            $show_html_default, $sort, $has_unsafe_images, $passed_ent_id;
 322
 323     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
 324         $view_unsafe_images = false;
 325     }
 326
 327     $body = '';
 328     $urlmailbox = urlencode($mailbox);
 329     $body_message = getEntity($message, $ent_num);
 330     if (($body_message->header->type0 == 'text') ||
 331         ($body_message->header->type0 == 'rfc822')) {
 332         $body = mime_fetch_body ($imap_stream, $id, $ent_num);
 333         $body = decodeBody($body, $body_message->header->encoding);
 334         $hookResults = do_hook("message_body", $body);
 335         $body = $hookResults[1];
 336
 337         /* If there are other types that shouldn't be formatted, add
 338          * them here.
 339          */
 340
 341         if ($body_message->header->type1 == 'html') {
 342             if ($show_html_default <> 1) {
 343                 $entity_conv = array('&nbsp;' => ' ',
 344                                      '<p>'    => "\n",
 345                                      '<br>'   => "\n",
 346                                      '<P>'    => "\n",
 347                                      '<BR>'   => "\n",
 348                                      '&gt;'   => '>',
 349                                      '&lt;'   => '<');
 350                 $body = strtr($body, $entity_conv);
 351                 $body = strip_tags($body);
 352                 $body = trim($body);
 353                 translateText($body, $wrap_at,
 354                               $body_message->header->getParameter('charset'));
 355             } else {
 356                 $body = magicHTML($body, $id, $message, $mailbox);
 357             }
 358         } else {
 359             translateText($body, $wrap_at,
 360                           $body_message->header->getParameter('charset'));
 361         }
 362         $link = 'read_body.php?passed_id=' . $id . '&amp;ent_id='.$ent_num.
 363                  '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
 364                  '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 365         if (isset($passed_ent_id)) {
 366             $link .= '&amp;passed_ent_id='.$passed_ent_id;
 367         }
 368         if ($view_unsafe_images) {
 369             $text = _("Hide Unsafe Images");
 370         } else {
 371             if (isset($has_unsafe_images) && $has_unsafe_images) {
 372                 $link .= '&amp;view_unsafe_images=1';
 373                 $text = _("View Unsafe Images");
 374             } else {
 375                 $text = '';
 376             }
 377         }
 378         $body .= '<center><small><a href="'.$link.'">'.$text.
 379                  '</a></small></center><br>' . "\n";
 380     }
 381     return $body;
 382 }
 383
 384
 385 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 386     global $where, $what, $startMessage, $color;
 387     static $ShownHTML = 0;
 388
 389     $att_ar = $message->getAttachments($exclude_id);
 390
 391     if (!count($att_ar)) return '';
 392
 393     $attachments = '';
 394
 395     $urlMailbox = urlencode($mailbox);
 396
 397     foreach ($att_ar as $att) {
 398         $ent = urldecode($att->entity_id);
 399         $header = $att->header;
 400         $type0 = strtolower($header->type0);
 401         $type1 = strtolower($header->type1);
 402         $name = '';
 403         $links['download link']['text'] = _("download");
 404         $links['download link']['href'] =
 405                 "../src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 406         $ImageURL = '';
 407         if ($type0 =='message' && $type1 == 'rfc822') {
 408             $default_page = '../src/read_body.php';
 409             $rfc822_header = $att->rfc822_header;
 410             $filename = $rfc822_header->subject;
 411             if (trim( $filename ) == '') {
 412                 $filename = 'untitled-[' . $ent . ']' ;
 413             }
 414             $from_o = $rfc822_header->from;
 415             if (is_object($from_o)) {
 416                 $from_name = $from_o->getAddress(false);
 417             } else {
 418                 $from_name = _("Unknown sender");
 419             }
 420             $from_name = decodeHeader(($from_name));
 421             $description = $from_name;
 422         } else {
 423             $default_page = '../src/download.php';
 424             if (is_object($header->disposition)) {
 425                 $filename = $header->disposition->getProperty('filename');
 426                 if (trim($filename) == '') {
 427                     $name = decodeHeader($header->disposition->getProperty('name'));
 428                     if (trim($name) == '') {
 429                         $name = $header->getParameter('name');
 430                         if(trim($name) == '') {
 431                             if (trim( $header->id ) == '') {
 432                                 $filename = 'untitled-[' . $ent . ']' ;
 433                             } else {
 434                                 $filename = 'cid: ' . $header->id;
 435                             }
 436                         } else {
 437                             $filename = $name;
 438                         }
 439                     } else {
 440                         $filename = $name;
 441                     }
 442                 }
 443             } else {
 444                 $filename = $header->getParameter('name');
 445                 if (!trim($filename)) {
 446                     if (trim( $header->id ) == '') {
 447                         $filename = 'untitled-[' . $ent . ']' ;
 448                     } else {
 449                         $filename = 'cid: ' . $header->id;
 450                     }
 451                 }
 452             }
 453             if ($header->description) {
 454                 $description = decodeHeader($header->description);
 455             } else {
 456                 $description = '';
 457             }
 458         }
 459
 460         $display_filename = $filename;
 461         if (isset($passed_ent_id)) {
 462             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 463         } else {
 464             $passed_ent_id_link = '';
 465         }
 466         $defaultlink = $default_page . "?startMessage=$startMessage"
 467                      . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 468                      . '&amp;ent_id='.$ent.$passed_ent_id_link.'&amp;absolute_dl=true';
 469         if ($where && $what) {
 470            $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 471         }
 472         /* This executes the attachment hook with a specific MIME-type.
 473          * If that doesn't have results, it tries if there's a rule
 474          * for a more generic type.
 475          */
 476         $hookresults = do_hook("attachment $type0/$type1", $links,
 477                                $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 478                                $display_filename, $where, $what);
 479         if(count($hookresults[1]) <= 1) {
 480             $hookresults = do_hook("attachment $type0/*", $links,
 481                                    $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 482                                    $display_filename, $where, $what);
 483         }
 484
 485         $links = $hookresults[1];
 486         $defaultlink = $hookresults[6];
 487
 488         $attachments .= '<TR><TD>' .
 489                         '<A HREF="'.$defaultlink.'">'.decodeHeader($display_filename).'</A>&nbsp;</TD>' .
 490                         '<TD><SMALL><b>' . show_readable_size($header->size) .
 491                         '</b>&nbsp;&nbsp;</small></TD>' .
 492                         "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
 493                         '<TD><SMALL>';
 494         $attachments .= '<b>' . $description . '</b>';
 495         $attachments .= '</SMALL></TD><TD><SMALL>&nbsp;';
 496
 497         $skipspaces = 1;
 498         foreach ($links as $val) {
 499             if ($skipspaces) {
 500                 $skipspaces = 0;
 501             } else {
 502                 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
 503             }
 504             $attachments .= '<a href="' . $val['href'] . '">' .  $val['text'] . '</a>';
 505         }
 506         unset($links);
 507         $attachments .= "</TD></TR>\n";
 508     }
 509     return $attachments;
 510 }
 511
 512 /* This function decodes the body depending on the encoding type. */
 513 function decodeBody($body, $encoding) {
 514     global $languages, $squirrelmail_language;
 515     global $show_html_default;
 516
 517     $body = str_replace("\r\n", "\n", $body);
 518     $encoding = strtolower($encoding);
 519
 520     if ($encoding == 'quoted-printable' ||
 521         $encoding == 'quoted_printable') {
 522         $body = quoted_printable_decode($body);
 523
 524         while (ereg("=\n", $body)) {
 525             $body = ereg_replace ("=\n", '', $body);
 526         }
 527
 528     } else if ($encoding == 'base64') {
 529         $body = base64_decode($body);
 530     }
 531
 532     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 533         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 534         $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
 535     }
 536
 537     // All other encodings are returned raw.
 538     return $body;
 539 }
 540
 541 /*
 542  * This functions decode strings that is encoded according to
 543  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 544  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 545  */
 546 function decodeHeader ($string, $utfencode=true,$htmlsave=true) {
 547     global $languages, $squirrelmail_language;
 548     if (is_array($string)) {
 549         $string = implode("\n", $string);
 550     }
 551
 552     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 553         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 554         $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
 555         // Do we need to return at this point?
 556         // return $string;
 557     }
 558     $i = 0;
 559     $aString = explode(' ',$string);
 560     foreach ($aString as $chunk) {
 561         $encoded = false;
 562         $aString[$i] = '';
 563         while (preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 564             //$aString[$i] .= $res[1];
 565             //echo "match ". $res[5] . "<br>";
 566             $encoding = ucfirst($res[3]);
 567             switch ($encoding)
 568             {
 569             case 'B':
 570                 $replace = base64_decode($res[4]);
 571                 $aString[$i] .= charset_decode($res[2],$replace);
 572                 break;
 573             case 'Q':
 574                 $replace = str_replace('_', ' ', $res[4]);
 575                 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 576                                     $replace);
 577                 /* Only encode into entities by default. Some places
 578                  * don't need the encoding, like the compose form.
 579                  */
 580                 if ($utfencode) {
 581                     $replace = charset_decode($res[2], $replace);
 582                 } else {
 583                     if ($htmlsave) {
 584                         $replace = htmlspecialchars($res[4]);
 585                     }
 586                 }
 587                 $aString[$i] .= $replace;
 588                 break;
 589             default:
 590                 break;
 591             }
 592             $chunk = $res[5];
 593             $encoded = true;
 594         }
 595         if (!$encoded && $htmlsave) {
 596             $aString[$i] = htmlspecialchars($chunk);
 597         } else {
 598             $aString[$i] .= $chunk;
 599         }
 600         ++$i;
 601     }
 602     if ($htmlsave) {
 603         return implode('&nbsp;',$aString);
 604     } else {
 605         return implode (' ',$aString);
 606     }
 607 }
 608
 609 /*
 610  * Encode a string according to RFC 1522 for use in headers if it
 611  * contains 8-bit characters or anything that looks like it should
 612  * be encoded.
 613  */
 614 function encodeHeader ($string) {
 615     global $default_charset, $languages, $squirrelmail_language;
 616
 617     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 618         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 619         return  $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
 620     }
 621
 622     // Encode only if the string contains 8-bit characters or =?
 623     $j = strlen($string);
 624     $l = strstr($string, '=?');         // Must be encoded ?
 625     $max_l = 75 - strlen($default_charset) - 7;
 626     $aRet = array();
 627     $ret = '';
 628     $enc = false;
 629     $cur_l = $iOffset = 0;
 630     for($i = 0; $i < $j; ++$i) {
 631         switch($string{$i}) {
 632             case '=':
 633                 $cur_l+=3;
 634                 if ($cur_l > $max_l) {
 635                     if ($enc) {
 636                         $aRet[] = "=?$default_charset?Q?$ret?=";
 637                         $enc = false;
 638                     } else {
 639                         $aRet[] = substr($string,$iOffset,$i-$iOffset+1);
 640                     }
 641                     $iOffset = $i+1;
 642                     $cur_l = 3;
 643                     $ret = '';
 644                 }
 645                 $ret .= '=3D';
 646                 break;
 647             case '?':
 648                 $cur_l+=3;
 649                 if ($cur_l > $max_l) {
 650                     if ($enc) {
 651                         $aRet[] = "=?$default_charset?Q?$ret?=";
 652                         $enc = false;
 653                     } else {
 654                         $aRet[] = substr($string,$iOffset,$i-$iOffset+1);
 655                     }
 656                     $iOffset = $i+1;
 657                     $cur_l = 3;
 658                     $ret = '';
 659                 }
 660                 $ret .= '=3F';
 661                 break;
 662             case '_':
 663                 $cur_l+=3;
 664                 if ($cur_l > $max_l) {
 665                     if ($enc) {
 666                         $aRet[] = "=?$default_charset?Q?$ret?=";
 667                         $enc = false;
 668                     } else {
 669                         $aRet[] = substr($string,$iOffset,$i-$iOffset+1);
 670                     }
 671                     $iOffset = $i+1;
 672                     $cur_l = 3;
 673                     $ret = '';
 674                 }
 675                 $ret .= '=5F';
 676                 break;
 677             case ' ':
 678                 $cur_l++;
 679                 if ($cur_l > $max_l) {
 680                     if ($enc) {
 681                         $aRet[] = "=?$default_charset?Q?$ret?=";
 682                         $enc = false;
 683                     } else {
 684                         $aRet[] = substr($string,$iOffset,$i-$iOffset+1);
 685                     }
 686                     $iOffset = $i+1;
 687                     $cur_l = 1;
 688                     $ret = '';
 689                 }
 690                 $ret .= '_';
 691                 break;
 692             default:
 693                 $k = ord($string{$i});
 694                 if ($k > 126) {
 695                     $enc = true;
 696                     $s = sprintf("=%02X", $k);
 697                     $cur_l += strlen($s);
 698                     if ($cur_l > $max_l) {
 699                         $aRet[] = "=?$default_charset?Q?$ret?=";
 700                         $cur_l = strlen($s);
 701                         $ret = '';
 702                         $l = false;
 703                         $enc = false;
 704                         $iOffset = $i+1;
 705                     }
 706                     $ret .= $s;
 707                     $l = TRUE;
 708                 } else {
 709                     $cur_l++;
 710                     if ($cur_l > $max_l) {
 711                         if ($enc) {
 712                            $aRet[] = "=?$default_charset?Q?$ret?=";
 713                            $enc = false;
 714                         } else {
 715                            $aRet[] = substr($ret,$iOffset,$i-$iOffset+1);
 716                         }
 717                         $iOffset = $i+1;
 718                         $l = false;
 719                         $cur_l = 1;
 720                         $ret = '';
 721                     }
 722                     $ret .= $string{$i};
 723                 }
 724                 break;
 725         }
 726     }
 727
 728     if ($enc) {
 729         $string = implode('',$aRet) . "=?$default_charset?Q?$ret?=";
 730     }
 731
 732     return $string;
 733 }
 734
 735 /* This function trys to locate the entity_id of a specific mime element */
 736 function find_ent_id($id, $message) {
 737     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
 738         if ($message->entities[$i]->header->type0 == 'multipart')  {
 739             $ret = find_ent_id($id, $message->entities[$i]);
 740         } else {
 741             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
 742 //                if (sq_check_save_extension($message->entities[$i])) {
 743                     return $message->entities[$i]->entity_id;
 744 //                }
 745             }
 746         }
 747     }
 748     return $ret;
 749 }
 750
 751 function sq_check_save_extension($message) {
 752     $filename = $message->getFilename();
 753     $ext = substr($filename, strrpos($filename,'.')+1);
 754     $save_extensions = array('jpg','jpeg','gif','png','bmp');
 755     return in_array($ext, $save_extensions);
 756 }
 757
 758
 759 /**
 760  ** HTMLFILTER ROUTINES
 761  */
 762
 763 /**
 764  * This function returns the final tag out of the tag name, an array
 765  * of attributes, and the type of the tag. This function is called by
 766  * sq_sanitize internally.
 767  *
 768  * @param  $tagname  the name of the tag.
 769  * @param  $attary   the array of attributes and their values
 770  * @param  $tagtype  The type of the tag (see in comments).
 771  * @return           a string with the final tag representation.
 772  */
 773 function sq_tagprint($tagname, $attary, $tagtype){
 774     $me = 'sq_tagprint';
 775
 776     if ($tagtype == 2){
 777         $fulltag = '</' . $tagname . '>';
 778     } else {
 779         $fulltag = '<' . $tagname;
 780         if (is_array($attary) && sizeof($attary)){
 781             $atts = Array();
 782             while (list($attname, $attvalue) = each($attary)){
 783                 array_push($atts, "$attname=$attvalue");
 784             }
 785             $fulltag .= ' ' . join(" ", $atts);
 786         }
 787         if ($tagtype == 3){
 788             $fulltag .= ' /';
 789         }
 790         $fulltag .= '>';
 791     }
 792     return $fulltag;
 793 }
 794
 795 /**
 796  * A small helper function to use with array_walk. Modifies a by-ref
 797  * value and makes it lowercase.
 798  *
 799  * @param  $val a value passed by-ref.
 800  * @return      void since it modifies a by-ref value.
 801  */
 802 function sq_casenormalize(&$val){
 803     $val = strtolower($val);
 804 }
 805
 806 /**
 807  * This function skips any whitespace from the current position within
 808  * a string and to the next non-whitespace value.
 809  *
 810  * @param  $body   the string
 811  * @param  $offset the offset within the string where we should start
 812  *                 looking for the next non-whitespace character.
 813  * @return         the location within the $body where the next
 814  *                 non-whitespace char is located.
 815  */
 816 function sq_skipspace($body, $offset){
 817     $me = 'sq_skipspace';
 818     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
 819     if (sizeof($matches{1})){
 820         $count = strlen($matches{1});
 821         $offset += $count;
 822     }
 823     return $offset;
 824 }
 825
 826 /**
 827  * This function looks for the next character within a string.  It's
 828  * really just a glorified "strpos", except it catches if failures
 829  * nicely.
 830  *
 831  * @param  $body   The string to look for needle in.
 832  * @param  $offset Start looking from this position.
 833  * @param  $needle The character/string to look for.
 834  * @return         location of the next occurance of the needle, or
 835  *                 strlen($body) if needle wasn't found.
 836  */
 837 function sq_findnxstr($body, $offset, $needle){
 838     $me  = 'sq_findnxstr';
 839     $pos = strpos($body, $needle, $offset);
 840     if ($pos === FALSE){
 841         $pos = strlen($body);
 842     }
 843     return $pos;
 844 }
 845
 846 /**
 847  * This function takes a PCRE-style regexp and tries to match it
 848  * within the string.
 849  *
 850  * @param  $body   The string to look for needle in.
 851  * @param  $offset Start looking from here.
 852  * @param  $reg    A PCRE-style regex to match.
 853  * @return         Returns a false if no matches found, or an array
 854  *                 with the following members:
 855  *                 - integer with the location of the match within $body
 856  *                 - string with whatever content between offset and the match
 857  *                 - string with whatever it is we matched
 858  */
 859 function sq_findnxreg($body, $offset, $reg){
 860     $me = 'sq_findnxreg';
 861     $matches = Array();
 862     $retarr = Array();
 863     preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
 864     if (!$matches{0}){
 865         $retarr = false;
 866     } else {
 867         $retarr{0} = $offset + strlen($matches{1});
 868         $retarr{1} = $matches{1};
 869         $retarr{2} = $matches{2};
 870     }
 871     return $retarr;
 872 }
 873
 874 /**
 875  * This function looks for the next tag.
 876  *
 877  * @param  $body   String where to look for the next tag.
 878  * @param  $offset Start looking from here.
 879  * @return         false if no more tags exist in the body, or
 880  *                 an array with the following members:
 881  *                 - string with the name of the tag
 882  *                 - array with attributes and their values
 883  *                 - integer with tag type (1, 2, or 3)
 884  *                 - integer where the tag starts (starting "<")
 885  *                 - integer where the tag ends (ending ">")
 886  *                 first three members will be false, if the tag is invalid.
 887  */
 888 function sq_getnxtag($body, $offset){
 889     $me = 'sq_getnxtag';
 890     if ($offset > strlen($body)){
 891         return false;
 892     }
 893     $lt = sq_findnxstr($body, $offset, "<");
 894     if ($lt == strlen($body)){
 895         return false;
 896     }
 897     /**
 898      * We are here:
 899      * blah blah <tag attribute="value">
 900      * \---------^
 901      */
 902     $pos = sq_skipspace($body, $lt+1);
 903     if ($pos >= strlen($body)){
 904         return Array(false, false, false, $lt, strlen($body));
 905     }
 906     /**
 907      * There are 3 kinds of tags:
 908      * 1. Opening tag, e.g.:
 909      *    <a href="blah">
 910      * 2. Closing tag, e.g.:
 911      *    </a>
 912      * 3. XHTML-style content-less tag, e.g.:
 913      *    <img src="blah"/>
 914      */
 915     $tagtype = false;
 916     switch (substr($body, $pos, 1)){
 917         case '/':
 918             $tagtype = 2;
 919             $pos++;
 920             break;
 921         case '!':
 922             /**
 923              * A comment or an SGML declaration.
 924              */
 925             if (substr($body, $pos+1, 2) == "--"){
 926                 $gt = strpos($body, "-->", $pos);
 927                 if ($gt === false){
 928                     $gt = strlen($body);
 929                 } else {
 930                     $gt += 2;
 931                 }
 932                 return Array(false, false, false, $lt, $gt);
 933             } else {
 934                 $gt = sq_findnxstr($body, $pos, ">");
 935                 return Array(false, false, false, $lt, $gt);
 936             }
 937             break;
 938         default:
 939             /**
 940              * Assume tagtype 1 for now. If it's type 3, we'll switch values
 941              * later.
 942              */
 943             $tagtype = 1;
 944             break;
 945     }
 946
 947     $tag_start = $pos;
 948     $tagname = '';
 949     /**
 950      * Look for next [\W-_], which will indicate the end of the tag name.
 951      */
 952     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
 953     if ($regary == false){
 954         return Array(false, false, false, $lt, strlen($body));
 955     }
 956     list($pos, $tagname, $match) = $regary;
 957     $tagname = strtolower($tagname);
 958
 959     /**
 960      * $match can be either of these:
 961      * '>'  indicating the end of the tag entirely.
 962      * '\s' indicating the end of the tag name.
 963      * '/'  indicating that this is type-3 xhtml tag.
 964      *
 965      * Whatever else we find there indicates an invalid tag.
 966      */
 967     switch ($match){
 968         case '/':
 969             /**
 970              * This is an xhtml-style tag with a closing / at the
 971              * end, like so: <img src="blah"/>. Check if it's followed
 972              * by the closing bracket. If not, then this tag is invalid
 973              */
 974             if (substr($body, $pos, 2) == "/>"){
 975                 $pos++;
 976                 $tagtype = 3;
 977             } else {
 978                 $gt = sq_findnxstr($body, $pos, ">");
 979                 $retary = Array(false, false, false, $lt, $gt);
 980                 return $retary;
 981             }
 982         case '>':
 983             return Array($tagname, false, $tagtype, $lt, $pos);
 984             break;
 985         default:
 986             /**
 987              * Check if it's whitespace
 988              */
 989             if (!preg_match('/\s/', $match)){
 990                 /**
 991                  * This is an invalid tag! Look for the next closing ">".
 992                  */
 993                 $gt = sq_findnxstr($body, $offset, ">");
 994                 return Array(false, false, false, $lt, $gt);
 995             }
 996             break;
 997     }
 998
 999     /**
1000      * At this point we're here:
1001      * <tagname  attribute='blah'>
1002      * \-------^
1003      *
1004      * At this point we loop in order to find all attributes.
1005      */
1006     $attname = '';
1007     $atttype = false;
1008     $attary = Array();
1009
1010     while ($pos <= strlen($body)){
1011         $pos = sq_skipspace($body, $pos);
1012         if ($pos == strlen($body)){
1013             /**
1014              * Non-closed tag.
1015              */
1016             return Array(false, false, false, $lt, $pos);
1017         }
1018         /**
1019          * See if we arrived at a ">" or "/>", which means that we reached
1020          * the end of the tag.
1021          */
1022         $matches = Array();
1023         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1024             /**
1025              * Yep. So we did.
1026              */
1027             $pos += strlen($matches{1});
1028             if ($matches{2} == "/>"){
1029                 $tagtype = 3;
1030                 $pos++;
1031             }
1032             return Array($tagname, $attary, $tagtype, $lt, $pos);
1033         }
1034
1035         /**
1036          * There are several types of attributes, with optional
1037          * [:space:] between members.
1038          * Type 1:
1039          *   attrname[:space:]=[:space:]'CDATA'
1040          * Type 2:
1041          *   attrname[:space:]=[:space:]"CDATA"
1042          * Type 3:
1043          *   attr[:space:]=[:space:]CDATA
1044          * Type 4:
1045          *   attrname
1046          *
1047          * We leave types 1 and 2 the same, type 3 we check for
1048          * '"' and convert to "&quot" if needed, then wrap in
1049          * double quotes. Type 4 we convert into:
1050          * attrname="yes".
1051          */
1052         $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1053         if ($regary == false){
1054             /**
1055              * Looks like body ended before the end of tag.
1056              */
1057             return Array(false, false, false, $lt, strlen($body));
1058         }
1059         list($pos, $attname, $match) = $regary;
1060         $attname = strtolower($attname);
1061         /**
1062          * We arrived at the end of attribute name. Several things possible
1063          * here:
1064          * '>'  means the end of the tag and this is attribute type 4
1065          * '/'  if followed by '>' means the same thing as above
1066          * '\s' means a lot of things -- look what it's followed by.
1067          *      anything else means the attribute is invalid.
1068          */
1069         switch($match){
1070             case '/':
1071                 /**
1072                  * This is an xhtml-style tag with a closing / at the
1073                  * end, like so: <img src="blah"/>. Check if it's followed
1074                  * by the closing bracket. If not, then this tag is invalid
1075                  */
1076                 if (substr($body, $pos, 2) == "/>"){
1077                     $pos++;
1078                     $tagtype = 3;
1079                 } else {
1080                     $gt = sq_findnxstr($body, $pos, ">");
1081                     $retary = Array(false, false, false, $lt, $gt);
1082                     return $retary;
1083                 }
1084             case '>':
1085                 $attary{$attname} = '"yes"';
1086                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1087                 break;
1088             default:
1089                 /**
1090                  * Skip whitespace and see what we arrive at.
1091                  */
1092                 $pos = sq_skipspace($body, $pos);
1093                 $char = substr($body, $pos, 1);
1094                 /**
1095                  * Two things are valid here:
1096                  * '=' means this is attribute type 1 2 or 3.
1097                  * \w means this was attribute type 4.
1098                  * anything else we ignore and re-loop. End of tag and
1099                  * invalid stuff will be caught by our checks at the beginning
1100                  * of the loop.
1101                  */
1102                 if ($char == "="){
1103                     $pos++;
1104                     $pos = sq_skipspace($body, $pos);
1105                     /**
1106                      * Here are 3 possibilities:
1107                      * "'"  attribute type 1
1108                      * '"'  attribute type 2
1109                      * everything else is the content of tag type 3
1110                      */
1111                     $quot = substr($body, $pos, 1);
1112                     if ($quot == "'"){
1113                         $regary = sq_findnxreg($body, $pos+1, "\'");
1114                         if ($regary == false){
1115                             return Array(false, false, false, $lt, strlen($body));
1116                         }
1117                         list($pos, $attval, $match) = $regary;
1118                         $pos++;
1119                         $attary{$attname} = "'" . $attval . "'";
1120                     } else if ($quot == '"'){
1121                         $regary = sq_findnxreg($body, $pos+1, '\"');
1122                         if ($regary == false){
1123                             return Array(false, false, false, $lt, strlen($body));
1124                         }
1125                         list($pos, $attval, $match) = $regary;
1126                         $pos++;
1127                         $attary{$attname} = '"' . $attval . '"';
1128                     } else {
1129                         /**
1130                          * These are hateful. Look for \s, or >.
1131                          */
1132                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1133                         if ($regary == false){
1134                             return Array(false, false, false, $lt, strlen($body));
1135                         }
1136                         list($pos, $attval, $match) = $regary;
1137                         /**
1138                          * If it's ">" it will be caught at the top.
1139                          */
1140                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1141                         $attary{$attname} = '"' . $attval . '"';
1142                     }
1143                 } else if (preg_match("|[\w/>]|", $char)) {
1144                     /**
1145                      * That was attribute type 4.
1146                      */
1147                     $attary{$attname} = '"yes"';
1148                 } else {
1149                     /**
1150                      * An illegal character. Find next '>' and return.
1151                      */
1152                     $gt = sq_findnxstr($body, $pos, ">");
1153                     return Array(false, false, false, $lt, $gt);
1154                 }
1155                 break;
1156         }
1157     }
1158     /**
1159      * The fact that we got here indicates that the tag end was never
1160      * found. Return invalid tag indication so it gets stripped.
1161      */
1162     return Array(false, false, false, $lt, strlen($body));
1163 }
1164
1165 /**
1166  * This function checks attribute values for entity-encoded values
1167  * and returns them translated into 8-bit strings so we can run
1168  * checks on them.
1169  *
1170  * @param  $attvalue A string to run entity check against.
1171  * @return           Translated value.
1172  */
1173 function sq_deent($attvalue){
1174     $me = 'sq_deent';
1175     /**
1176      * See if we have to run the checks first. All entities must start
1177      * with "&".
1178      */
1179     if (strpos($attvalue, "&") === false){
1180         return $attvalue;
1181     }
1182     /**
1183      * Check named entities first.
1184      */
1185     $trans = get_html_translation_table(HTML_ENTITIES);
1186     /**
1187      * Leave &quot; in, as it can mess us up.
1188      */
1189     $trans = array_flip($trans);
1190     unset($trans{"&quot;"});
1191     while (list($ent, $val) = each($trans)){
1192         $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1193     }
1194     /**
1195      * Now translate numbered entities from 1 to 255 if needed.
1196      */
1197     if (strpos($attvalue, "#") !== false){
1198         $omit = Array(34, 39);
1199         for ($asc=1; $asc<256; $asc++){
1200             if (!in_array($asc, $omit)){
1201                 $chr = chr($asc);
1202                 $attvalue = preg_replace("/\&#0*$asc;*(\D)/si", "$chr\\1",
1203                                          $attvalue);
1204                 $attvalue = preg_replace("/\&#x0*".dechex($asc).";*(\W)/si",
1205                                          "$chr\\1", $attvalue);
1206             }
1207         }
1208     }
1209     return $attvalue;
1210 }
1211
1212 /**
1213  * This function runs various checks against the attributes.
1214  *
1215  * @param  $tagname         String with the name of the tag.
1216  * @param  $attary          Array with all tag attributes.
1217  * @param  $rm_attnames     See description for sq_sanitize
1218  * @param  $bad_attvals     See description for sq_sanitize
1219  * @param  $add_attr_to_tag See description for sq_sanitize
1220  * @param  $message         message object
1221  * @param  $id              message id
1222  * @return                  Array with modified attributes.
1223  */
1224 function sq_fixatts($tagname,
1225                     $attary,
1226                     $rm_attnames,
1227                     $bad_attvals,
1228                     $add_attr_to_tag,
1229                     $message,
1230                     $id,
1231                     $mailbox
1232                     ){
1233     $me = 'sq_fixatts';
1234     while (list($attname, $attvalue) = each($attary)){
1235         /**
1236          * See if this attribute should be removed.
1237          */
1238         foreach ($rm_attnames as $matchtag=>$matchattrs){
1239             if (preg_match($matchtag, $tagname)){
1240                 foreach ($matchattrs as $matchattr){
1241                     if (preg_match($matchattr, $attname)){
1242                         unset($attary{$attname});
1243                         continue;
1244                     }
1245                 }
1246             }
1247         }
1248         /**
1249          * Remove any entities.
1250          */
1251         $attvalue = sq_deent($attvalue);
1252
1253         /**
1254          * Now let's run checks on the attvalues.
1255          * I don't expect anyone to comprehend this. If you do,
1256          * get in touch with me so I can drive to where you live and
1257          * shake your hand personally. :)
1258          */
1259         foreach ($bad_attvals as $matchtag=>$matchattrs){
1260             if (preg_match($matchtag, $tagname)){
1261                 foreach ($matchattrs as $matchattr=>$valary){
1262                     if (preg_match($matchattr, $attname)){
1263                         /**
1264                          * There are two arrays in valary.
1265                          * First is matches.
1266                          * Second one is replacements
1267                          */
1268                         list($valmatch, $valrepl) = $valary;
1269                         $newvalue =
1270                             preg_replace($valmatch, $valrepl, $attvalue);
1271                         if ($newvalue != $attvalue){
1272                             $attary{$attname} = $newvalue;
1273                         }
1274                     }
1275                 }
1276             }
1277         }
1278         /**
1279          * Turn cid: urls into http-friendly ones.
1280          */
1281         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1282             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1283         }
1284     }
1285     /**
1286      * See if we need to append any attributes to this tag.
1287      */
1288     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1289         if (preg_match($matchtag, $tagname)){
1290             $attary = array_merge($attary, $addattary);
1291         }
1292     }
1293     return $attary;
1294 }
1295
1296 /**
1297  * This function edits the style definition to make them friendly and
1298  * usable in squirrelmail.
1299  *
1300  * @param  $message  the message object
1301  * @param  $id       the message id
1302  * @param  $content  a string with whatever is between <style> and </style>
1303  * @return           a string with edited content.
1304  */
1305 function sq_fixstyle($message, $id, $content){
1306     global $view_unsafe_images;
1307     $me = 'sq_fixstyle';
1308     /**
1309      * First look for general BODY style declaration, which would be
1310      * like so:
1311      * body {background: blah-blah}
1312      * and change it to .bodyclass so we can just assign it to a <div>
1313      */
1314     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1315     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1316     /**
1317      * Fix url('blah') declarations.
1318      */
1319     $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1320                             "url(\\1$secremoveimg\\2)", $content);
1321     /**
1322      * Fix url('https*://.*) declarations but only if $view_unsafe_images
1323      * is false.
1324      */
1325     if (!$view_unsafe_images){
1326         $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1327                                 "url(\\1$secremoveimg\\2)", $content);
1328     }
1329
1330     /**
1331      * Fix urls that refer to cid:
1332      */
1333     while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1334                       $matches)){
1335         $cidurl = $matches{1};
1336         $httpurl = sq_cid2http($message, $id, $cidurl);
1337         $content = preg_replace("|url\($cidurl\)|si",
1338                                 "url($httpurl)", $content);
1339     }
1340
1341     /**
1342      * Fix stupid css declarations which lead to vulnerabilities
1343      * in IE.
1344      */
1345     $match   = Array('/expression/si',
1346                      '/behaviou*r/si',
1347                      '/binding/si');
1348     $replace = Array('idiocy', 'idiocy', 'idiocy');
1349     $content = preg_replace($match, $replace, $content);
1350     return $content;
1351 }
1352
1353 /**
1354  * This function converts cid: url's into the ones that can be viewed in
1355  * the browser.
1356  *
1357  * @param  $message  the message object
1358  * @param  $id       the message id
1359  * @param  $cidurl   the cid: url.
1360  * @return           a string with a http-friendly url
1361  */
1362 function sq_cid2http($message, $id, $cidurl, $mailbox){
1363     /**
1364      * Get rid of quotes.
1365      */
1366     $quotchar = substr($cidurl, 0, 1);
1367     $cidurl = str_replace($quotchar, "", $cidurl);
1368     $cidurl = substr(trim($cidurl), 4);
1369     $linkurl = find_ent_id($cidurl, $message);
1370     /* in case of non-save cid links $httpurl should be replaced by a sort of
1371        unsave link image */
1372     $httpurl = '';
1373     if ($linkurl) {
1374         $httpurl = $quotchar . '../src/download.php?absolute_dl=true&amp;' .
1375                    "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1376                    '&amp;ent_id=' . $linkurl . $quotchar;
1377     }
1378     return $httpurl;
1379 }
1380
1381 /**
1382  * This function changes the <body> tag into a <div> tag since we
1383  * can't really have a body-within-body.
1384  *
1385  * @param  $attary  an array of attributes and values of <body>
1386  * @return          a modified array of attributes to be set for <div>
1387  */
1388 function sq_body2div($attary){
1389     $me = 'sq_body2div';
1390     $divattary = Array('class' => "'bodyclass'");
1391     $bgcolor = '#ffffff';
1392     $text = '#000000';
1393     $styledef = '';
1394     if (is_array($attary) && sizeof($attary) > 0){
1395         foreach ($attary as $attname=>$attvalue){
1396             $quotchar = substr($attvalue, 0, 1);
1397             $attvalue = str_replace($quotchar, "", $attvalue);
1398             switch ($attname){
1399                 case 'background':
1400                     $styledef .= "background-image: url('$attvalue'); ";
1401                     break;
1402                 case 'bgcolor':
1403                     $styledef .= "background-color: $attvalue; ";
1404                     break;
1405                 case 'text':
1406                     $styledef .= "color: $attvalue; ";
1407                     break;
1408             }
1409         }
1410         if (strlen($styledef) > 0){
1411             $divattary{"style"} = "\"$styledef\"";
1412         }
1413     }
1414     return $divattary;
1415 }
1416
1417 /**
1418  * This is the main function and the one you should actually be calling.
1419  * There are several variables you should be aware of an which need
1420  * special description.
1421  *
1422  * Since the description is quite lengthy, see it here:
1423  * http://www.mricon.com/html/phpfilter.html
1424  *
1425  * @param $body                 the string with HTML you wish to filter
1426  * @param $tag_list             see description above
1427  * @param $rm_tags_with_content see description above
1428  * @param $self_closing_tags    see description above
1429  * @param $force_tag_closing    see description above
1430  * @param $rm_attnames          see description above
1431  * @param $bad_attvals          see description above
1432  * @param $add_attr_to_tag      see description above
1433  * @param $message              message object
1434  * @param $id                   message id
1435  * @return                      sanitized html safe to show on your pages.
1436  */
1437 function sq_sanitize($body,
1438                      $tag_list,
1439                      $rm_tags_with_content,
1440                      $self_closing_tags,
1441                      $force_tag_closing,
1442                      $rm_attnames,
1443                      $bad_attvals,
1444                      $add_attr_to_tag,
1445                      $message,
1446                      $id,
1447                      $mailbox
1448                      ){
1449     $me = 'sq_sanitize';
1450     /**
1451      * Normalize rm_tags and rm_tags_with_content.
1452      */
1453     @array_walk($rm_tags, 'sq_casenormalize');
1454     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1455     @array_walk($self_closing_tags, 'sq_casenormalize');
1456     /**
1457      * See if tag_list is of tags to remove or tags to allow.
1458      * false  means remove these tags
1459      * true   means allow these tags
1460      */
1461     $rm_tags = array_shift($tag_list);
1462     $curpos = 0;
1463     $open_tags = Array();
1464     $trusted = "<!-- begin sanitized html -->\n";
1465     $skip_content = false;
1466     /**
1467      * Take care of netscape's stupid javascript entities like
1468      * &{alert('boo')};
1469      */
1470     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1471
1472     while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1473         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1474         $free_content = substr($body, $curpos, $lt-$curpos);
1475         /**
1476          * Take care of <style>
1477          */
1478         if ($tagname == "style" && $tagtype == 2){
1479             /**
1480              * This is a closing </style>. Edit the
1481              * content before we apply it.
1482              */
1483             $free_content = sq_fixstyle($message, $id, $free_content);
1484         }
1485         if ($skip_content == false){
1486             $trusted .= $free_content;
1487         }
1488         if ($tagname != FALSE){
1489             if ($tagtype == 2){
1490                 if ($skip_content == $tagname){
1491                     /**
1492                      * Got to the end of tag we needed to remove.
1493                      */
1494                     $tagname = false;
1495                     $skip_content = false;
1496                 } else {
1497                     if ($skip_content == false){
1498                         if ($tagname == "body"){
1499                             $tagname = "div";
1500                         } else {
1501                             if (isset($open_tags{$tagname}) &&
1502                                 $open_tags{$tagname} > 0){
1503                                 $open_tags{$tagname}--;
1504                             } else {
1505                                 $tagname = false;
1506                             }
1507                         }
1508                     }
1509                 }
1510             } else {
1511                 /**
1512                  * $rm_tags_with_content
1513                  */
1514                 if ($skip_content == false){
1515                     /**
1516                      * See if this is a self-closing type and change
1517                      * tagtype appropriately.
1518                      */
1519                     if ($tagtype == 1
1520                         && in_array($tagname, $self_closing_tags)){
1521                         $tagtype=3;
1522                     }
1523                     /**
1524                      * See if we should skip this tag and any content
1525                      * inside it.
1526                      */
1527                     if ($tagtype == 1 &&
1528                         in_array($tagname, $rm_tags_with_content)){
1529                         $skip_content = $tagname;
1530                     } else {
1531                         if (($rm_tags == false
1532                              && in_array($tagname, $tag_list)) ||
1533                             ($rm_tags == true &&
1534                              !in_array($tagname, $tag_list))){
1535                             $tagname = false;
1536                         } else {
1537                             if ($tagtype == 1){
1538                                 if (isset($open_tags{$tagname})){
1539                                     $open_tags{$tagname}++;
1540                                 } else {
1541                                     $open_tags{$tagname}=1;
1542                                 }
1543                             }
1544                             /**
1545                              * This is where we run other checks.
1546                              */
1547                             if (is_array($attary) && sizeof($attary) > 0){
1548                                 $attary = sq_fixatts($tagname,
1549                                                      $attary,
1550                                                      $rm_attnames,
1551                                                      $bad_attvals,
1552                                                      $add_attr_to_tag,
1553                                                      $message,
1554                                                      $id,
1555                                                      $mailbox
1556                                                      );
1557                             }
1558                             /**
1559                              * Convert body into div.
1560                              */
1561                             if ($tagname == "body"){
1562                                 $tagname = "div";
1563                                 $attary = sq_body2div($attary, $message, $id);
1564                             }
1565                         }
1566                     }
1567                 }
1568             }
1569             if ($tagname != false && $skip_content == false){
1570                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1571             }
1572         }
1573         $curpos = $gt+1;
1574     }
1575     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1576     if ($force_tag_closing == true){
1577         foreach ($open_tags as $tagname=>$opentimes){
1578             while ($opentimes > 0){
1579                 $trusted .= '</' . $tagname . '>';
1580                 $opentimes--;
1581             }
1582         }
1583         $trusted .= "\n";
1584     }
1585     $trusted .= "<!-- end sanitized html -->\n";
1586     return $trusted;
1587 }
1588
1589 /**
1590  * This is a wrapper function to call html sanitizing routines.
1591  *
1592  * @param  $body  the body of the message
1593  * @param  $id    the id of the message
1594  * @return        a string with html safe to display in the browser.
1595  */
1596 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1597     global $attachment_common_show_images, $view_unsafe_images,
1598            $has_unsafe_images;
1599     /**
1600      * Don't display attached images in HTML mode.
1601      */
1602     $attachment_common_show_images = false;
1603     $tag_list = Array(
1604                       false,
1605                       "object",
1606                       "meta",
1607                       "html",
1608                       "head",
1609                       "base",
1610                       "link",
1611                       "frame",
1612                       "iframe"
1613                       );
1614
1615     $rm_tags_with_content = Array(
1616                                   "script",
1617                                   "applet",
1618                                   "embed",
1619                                   "title"
1620                                   );
1621
1622     $self_closing_tags =  Array(
1623                                 "img",
1624                                 "br",
1625                                 "hr",
1626                                 "input"
1627                                 );
1628
1629     $force_tag_closing = false;
1630
1631     $rm_attnames = Array(
1632                          "/.*/" =>
1633                          Array(
1634                                "/target/si",
1635                                "/^on.*/si",
1636                                "/^dynsrc/si",
1637                                "/^data.*/si"
1638                                )
1639                          );
1640
1641     $secremoveimg = "../images/" . _("sec_remove_eng.png");
1642     $bad_attvals = Array(
1643         "/.*/" =>
1644             Array(
1645                 "/^src|background/i" =>
1646                     Array(
1647                           Array(
1648                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1649                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1650                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1651                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1652                                 ),
1653                           Array(
1654                                 "\\1$secremoveimg\\2",
1655                                 "\\1$secremoveimg\\2",
1656                                 "\\1$secremoveimg\\2",
1657                                 "\\1$secremoveimg\\2"
1658                                 )
1659                         ),
1660                 "/^href|action/i" =>
1661                     Array(
1662                           Array(
1663                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1664                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1665                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1666                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1667                                 ),
1668                           Array(
1669                                 "\\1#\\2",
1670                                 "\\1#\\2",
1671                                 "\\1#\\2",
1672                                 "\\1#\\2"
1673                                 )
1674                         ),
1675                 "/^style/si" =>
1676                     Array(
1677                           Array(
1678                                 "/expression/si",
1679                                 "/binding/si",
1680                                 "/behaviou*r/si",
1681                                 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1682                                 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1683                                 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1684                                 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1685                                ),
1686                           Array(
1687                                 "idiocy",
1688                                 "idiocy",
1689                                 "idiocy",
1690                                 "url(\\1#\\2)",
1691                                 "url(\\1#\\2)",
1692                                 "url(\\1#\\2)",
1693                                 "url(\\1#\\2)"
1694                                )
1695                           )
1696                 )
1697         );
1698     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1699         $view_unsafe_images = false;
1700     }
1701     if (!$view_unsafe_images){
1702         /**
1703          * Remove any references to http/https if view_unsafe_images set
1704          * to false.
1705          */
1706          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1707                     '/^([\'\"])\s*https*:.*([\'\"])/si');
1708          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1709                     "\\1$secremoveimg\\2");
1710          array_push($bad_attvals{'/.*/'}{'/^style/si'}[0],
1711                     '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1712          array_push($bad_attvals{'/.*/'}{'/^style/si'}[1],
1713                     "url(\\1$secremoveimg\\2)");
1714     }
1715
1716     $add_attr_to_tag = Array(
1717                              "/^a$/si" => Array('target'=>'"_new"')
1718                              );
1719     $trusted = sq_sanitize($body,
1720                            $tag_list,
1721                            $rm_tags_with_content,
1722                            $self_closing_tags,
1723                            $force_tag_closing,
1724                            $rm_attnames,
1725                            $bad_attvals,
1726                            $add_attr_to_tag,
1727                            $message,
1728                            $id,
1729                            $mailbox
1730                            );
1731     if (preg_match("|$secremoveimg|si", $trusted)){
1732         $has_unsafe_images = true;
1733     }
1734     return $trusted;
1735 }
1736
1737 ?>