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 $uid_support;
  99     /* Do a bit of error correction.  If we couldn't find the entity id, just guess
 100      * that it is the first one.  That is usually the case anyway.
 101      */
 102
 103     if (!$ent_id) {
 104         $cmd = "FETCH $id BODY[]";
 105     } else {
 106         $cmd = "FETCH $id BODY[$ent_id]";
 107     }
 108
 109     $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, $uid_support);
 110     do {
 111         $topline = trim(array_shift($data));
 112     } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
 113
 114     $wholemessage = implode('', $data);
 115     if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
 116         $ret = substr($wholemessage, 0, $regs[1]);
 117         /* There is some information in the content info header that could be important
 118          * in order to parse html messages. Let's get them here.
 119          */
 120 //        if ($ret{0} == '<') {
 121 //            $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, $uid_support);
 122 //        }
 123     } else if (ereg('"([^"]*)"', $topline, $regs)) {
 124         $ret = $regs[1];
 125     } else {
 126         global $where, $what, $mailbox, $passed_id, $startMessage;
 127         $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
 128         if (isset($where) && isset($what)) {
 129             $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
 130         } else {
 131             $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 132         }
 133         $par .= '&amp;response=' . urlencode($response) .
 134                 '&amp;message='  . urlencode($message)  .
 135                 '&amp;topline='  . urlencode($topline);
 136
 137         echo   '<tt><br>' .
 138                '<table width="80%"><tr>' .
 139                '<tr><td colspan=2>' .
 140                _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
 141                '</td></tr>' .
 142                '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
 143                '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
 144                '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
 145                '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
 146                "</table><BR></tt></font><hr>";
 147
 148         $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, $uid_support);
 149         array_shift($data);
 150         $wholemessage = implode('', $data);
 151
 152         $ret = $wholemessage;
 153     }
 154     return $ret;
 155 }
 156
 157 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding) {
 158     global $uid_support;
 159
 160     /* Don't kill the connection if the browser is over a dialup
 161      * and it would take over 30 seconds to download it.
 162      * Don´t call set_time_limit in safe mode.
 163      */
 164
 165     if (!ini_get('safe_mode')) {
 166         set_time_limit(0);
 167     }
 168     /* in case of base64 encoded attachments, do not buffer them.
 169        Instead, echo the decoded attachment directly to screen */
 170     if (strtolower($encoding) == 'base64') {
 171         if (!$ent_id) {
 172            $query = "FETCH $id BODY[]";
 173         } else {
 174            $query = "FETCH $id BODY[$ent_id]";
 175         }
 176         sqimap_run_command($imap_stream,$query,true,$response,$message,$uid_support,'sqimap_base64_decode','php://stdout',true);
 177     } else {
 178        $body = mime_fetch_body ($imap_stream, $id, $ent_id);
 179        echo decodeBody($body, $encoding);
 180     }
 181
 182     /*
 183        TODO, use the same method for quoted printable.
 184        However, I assume that quoted printable attachments aren't that large
 185        so the performancegain / memory usage drop will be minimal.
 186        If we decide to add that then we need to adapt sqimap_fread because
 187        we need to split te result on \n and fread doesn't stop at \n. That
 188        means we also should provide $results from sqimap_fread (by ref) to
 189        te function and set $no_return to false. The $filter function for
 190        quoted printable should handle unsetting of $results.
 191     */
 192     /*
 193        TODO 2: find out how we write to the output stream php://stdout. fwrite
 194        doesn't work because 'php://stdout isn't a stream.
 195     */
 196
 197     return;
 198 /*
 199     fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
 200     $cnt = 0;
 201     $continue = true;
 202     $read = fgets ($imap_stream,8192);
 203
 204
 205     // This could be bad -- if the section has sqimap_session_id() . ' OK'
 206     // or similar, it will kill the download.
 207     while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
 208         if (trim($read) == ')==') {
 209             $read1 = $read;
 210             $read = fgets ($imap_stream,4096);
 211             if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
 212                 return;
 213             } else {
 214                 echo decodeBody($read1, $encoding) .
 215                      decodeBody($read, $encoding);
 216             }
 217         } else if ($cnt) {
 218             echo decodeBody($read, $encoding);
 219         }
 220         $read = fgets ($imap_stream,4096);
 221         $cnt++;
 222 //      break;
 223     }
 224 */
 225 }
 226
 227 /* -[ END MIME DECODING ]----------------------------------------------------------- */
 228
 229 /* This is here for debugging purposes.  It will print out a list
 230  * of all the entity IDs that are in the $message object.
 231  */
 232 function listEntities ($message) {
 233     if ($message) {
 234         echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br>';
 235         for ($i = 0; isset($message->entities[$i]); $i++) {
 236             echo "$i : ";
 237             $msg = listEntities($message->entities[$i]);
 238
 239             if ($msg) {
 240                 echo "return: ";
 241                 return $msg;
 242             }
 243         }
 244     }
 245 }
 246
 247 function getPriorityStr($priority) {
 248     $priority_level = substr($priority,0,1);
 249
 250     switch($priority_level) {
 251         /* Check for a higher then normal priority. */
 252         case '1':
 253         case '2':
 254             $priority_string = _("High");
 255             break;
 256
 257         /* Check for a lower then normal priority. */
 258         case '4':
 259         case '5':
 260             $priority_string = _("Low");
 261             break;
 262
 263         /* Check for a normal priority. */
 264         case '3':
 265         default:
 266             $priority_level = '3';
 267             $priority_string = _("Normal");
 268             break;
 269
 270     }
 271     return $priority_string;
 272 }
 273
 274 /* returns a $message object for a particular entity id */
 275 function getEntity ($message, $ent_id) {
 276     return $message->getEntity($ent_id);
 277 }
 278
 279 /* translateText
 280  * Extracted from strings.php 23/03/2002
 281  */
 282
 283 function translateText(&$body, $wrap_at, $charset) {
 284     global $where, $what;   /* from searching */
 285     global $color;          /* color theme */
 286
 287     require_once(SM_PATH . 'functions/url_parser.php');
 288
 289     $body_ary = explode("\n", $body);
 290     for ($i=0; $i < count($body_ary); $i++) {
 291         $line = $body_ary[$i];
 292         if (strlen($line) - 2 >= $wrap_at) {
 293             sqWordWrap($line, $wrap_at);
 294         }
 295         $line = charset_decode($charset, $line);
 296         $line = str_replace("\t", '        ', $line);
 297
 298         parseUrl ($line);
 299
 300         $quotes = 0;
 301         $pos = 0;
 302         $j = strlen($line);
 303
 304         while ($pos < $j) {
 305             if ($line[$pos] == ' ') {
 306                 $pos++;
 307             } else if (strpos($line, '&gt;', $pos) === $pos) {
 308                 $pos += 4;
 309                 $quotes++;
 310             } else {
 311                 break;
 312             }
 313         }
 314
 315         if ($quotes % 2) {
 316             if (!isset($color[13])) {
 317                 $color[13] = '#800000';
 318             }
 319             $line = '<font color="' . $color[13] . '">' . $line . '</font>';
 320         } elseif ($quotes) {
 321             if (!isset($color[14])) {
 322                 $color[14] = '#FF0000';
 323             }
 324             $line = '<font color="' . $color[14] . '">' . $line . '</font>';
 325         }
 326
 327         $body_ary[$i] = $line;
 328     }
 329     $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
 330 }
 331
 332 /* This returns a parsed string called $body. That string can then
 333  * be displayed as the actual message in the HTML. It contains
 334  * everything needed, including HTML Tags, Attachments at the
 335  * bottom, etc.
 336  */
 337 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
 338     /* This if statement checks for the entity to show as the
 339      * primary message. To add more of them, just put them in the
 340      * order that is their priority.
 341      */
 342     global $startMessage, $username, $key, $imapServerAddress, $imapPort,
 343            $show_html_default, $sort, $has_unsafe_images, $passed_ent_id;
 344     global $languages, $squirrelmail_language;
 345
 346     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
 347         $view_unsafe_images = false;
 348     }
 349
 350     $body = '';
 351     $urlmailbox = urlencode($mailbox);
 352     $body_message = getEntity($message, $ent_num);
 353     if (($body_message->header->type0 == 'text') ||
 354         ($body_message->header->type0 == 'rfc822')) {
 355         $body = mime_fetch_body ($imap_stream, $id, $ent_num);
 356         $body = decodeBody($body, $body_message->header->encoding);
 357
 358         if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 359             function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 360             if (mb_detect_encoding($body) != 'ASCII') {
 361                 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
 362             }
 363         }
 364         $hookResults = do_hook("message_body", $body);
 365         $body = $hookResults[1];
 366
 367         /* If there are other types that shouldn't be formatted, add
 368          * them here.
 369          */
 370
 371         if ($body_message->header->type1 == 'html') {
 372             if ($show_html_default <> 1) {
 373                 $entity_conv = array('&nbsp;' => ' ',
 374                                      '<p>'    => "\n",
 375                                      '<br>'   => "\n",
 376                                      '<P>'    => "\n",
 377                                      '<BR>'   => "\n",
 378                                      '&gt;'   => '>',
 379                                      '&lt;'   => '<');
 380                 $body = strtr($body, $entity_conv);
 381                 $body = strip_tags($body);
 382                 $body = trim($body);
 383                 translateText($body, $wrap_at,
 384                               $body_message->header->getParameter('charset'));
 385             } else {
 386                 $body = magicHTML($body, $id, $message, $mailbox);
 387             }
 388         } else {
 389             translateText($body, $wrap_at,
 390                           $body_message->header->getParameter('charset'));
 391         }
 392         $link = 'read_body.php?passed_id=' . $id . '&amp;ent_id='.$ent_num.
 393                 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
 394                 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 395         if (isset($passed_ent_id)) {
 396             $link .= '&amp;passed_ent_id='.$passed_ent_id;
 397         }
 398         if ($view_unsafe_images) {
 399             $text = _("Hide Unsafe Images");
 400         } else {
 401             if (isset($has_unsafe_images) && $has_unsafe_images) {
 402                 $link .= '&amp;view_unsafe_images=1';
 403                 $text = _("View Unsafe Images");
 404             } else {
 405                 $text = '';
 406             }
 407         }
 408         $body .= '<center><small><a href="'.$link.'">'.$text.
 409                  '</a></small></center><br>' . "\n";
 410     }
 411     return $body;
 412 }
 413
 414
 415 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 416     global $where, $what, $startMessage, $color;
 417     static $ShownHTML = 0;
 418
 419     $att_ar = $message->getAttachments($exclude_id);
 420
 421     if (!count($att_ar)) return '';
 422
 423     $attachments = '';
 424
 425     $urlMailbox = urlencode($mailbox);
 426
 427     foreach ($att_ar as $att) {
 428         $ent = $att->entity_id;
 429         $header = $att->header;
 430         $type0 = strtolower($header->type0);
 431         $type1 = strtolower($header->type1);
 432         $name = '';
 433         $links['download link']['text'] = _("download");
 434         $links['download link']['href'] = SM_PATH .
 435                 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 436         $ImageURL = '';
 437         if ($type0 =='message' && $type1 == 'rfc822') {
 438             $default_page = SM_PATH . 'src/read_body.php';
 439             $rfc822_header = $att->rfc822_header;
 440             $filename = $rfc822_header->subject;
 441             if (trim( $filename ) == '') {
 442                 $filename = 'untitled-[' . $ent . ']' ;
 443             }
 444             $from_o = $rfc822_header->from;
 445             if (is_object($from_o)) {
 446                 $from_name = $from_o->getAddress(false);
 447             } else {
 448                 $from_name = _("Unknown sender");
 449             }
 450             $from_name = decodeHeader(($from_name));
 451             $description = $from_name;
 452         } else {
 453             $default_page = SM_PATH . 'src/download.php';
 454             if (is_object($header->disposition)) {
 455                 $filename = $header->disposition->getProperty('filename');
 456                 if (trim($filename) == '') {
 457                     $name = decodeHeader($header->disposition->getProperty('name'));
 458                     if (trim($name) == '') {
 459                         $name = $header->getParameter('name');
 460                         if(trim($name) == '') {
 461                             if (trim( $header->id ) == '') {
 462                                 $filename = 'untitled-[' . $ent . ']' ;
 463                             } else {
 464                                 $filename = 'cid: ' . $header->id;
 465                             }
 466                         } else {
 467                             $filename = $name;
 468                         }
 469                     } else {
 470                         $filename = $name;
 471                     }
 472                 }
 473             } else {
 474                 $filename = $header->getParameter('name');
 475                 if (!trim($filename)) {
 476                     if (trim( $header->id ) == '') {
 477                         $filename = 'untitled-[' . $ent . ']' ;
 478                     } else {
 479                         $filename = 'cid: ' . $header->id;
 480                     }
 481                 }
 482             }
 483             if ($header->description) {
 484                 $description = decodeHeader($header->description);
 485             } else {
 486                 $description = '';
 487             }
 488         }
 489
 490         $display_filename = $filename;
 491         if (isset($passed_ent_id)) {
 492             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 493         } else {
 494             $passed_ent_id_link = '';
 495         }
 496         $defaultlink = $default_page . "?startMessage=$startMessage"
 497                      . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 498                      . '&amp;ent_id='.$ent.$passed_ent_id_link;
 499         if ($where && $what) {
 500            $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 501         }
 502         /* This executes the attachment hook with a specific MIME-type.
 503          * If that doesn't have results, it tries if there's a rule
 504          * for a more generic type.
 505          */
 506         $hookresults = do_hook("attachment $type0/$type1", $links,
 507                                $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 508                                $display_filename, $where, $what);
 509         if(count($hookresults[1]) <= 1) {
 510             $hookresults = do_hook("attachment $type0/*", $links,
 511                                    $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 512                                    $display_filename, $where, $what);
 513         }
 514
 515         $links = $hookresults[1];
 516         $defaultlink = $hookresults[6];
 517
 518         $attachments .= '<TR><TD>' .
 519                         '<A HREF="'.$defaultlink.'">'.decodeHeader($display_filename).'</A>&nbsp;</TD>' .
 520                         '<TD><SMALL><b>' . show_readable_size($header->size) .
 521                         '</b>&nbsp;&nbsp;</small></TD>' .
 522                         "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
 523                         '<TD><SMALL>';
 524         $attachments .= '<b>' . $description . '</b>';
 525         $attachments .= '</SMALL></TD><TD><SMALL>&nbsp;';
 526
 527         $skipspaces = 1;
 528         foreach ($links as $val) {
 529             if ($skipspaces) {
 530                 $skipspaces = 0;
 531             } else {
 532                 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
 533             }
 534             $attachments .= '<a href="' . $val['href'] . '">' .  $val['text'] . '</a>';
 535         }
 536         unset($links);
 537         $attachments .= "</TD></TR>\n";
 538     }
 539     return $attachments;
 540 }
 541
 542 function sqimap_base64_decode(&$string) {
 543     $string = base64_decode($string);
 544 }
 545
 546 /* This function decodes the body depending on the encoding type. */
 547 function decodeBody($body, $encoding) {
 548     global $show_html_default;
 549
 550     $body = str_replace("\r\n", "\n", $body);
 551     $encoding = strtolower($encoding);
 552
 553     if ($encoding == 'quoted-printable' ||
 554         $encoding == 'quoted_printable') {
 555         $body = quoted_printable_decode($body);
 556
 557         while (ereg("=\n", $body)) {
 558             $body = ereg_replace ("=\n", '', $body);
 559         }
 560
 561     } else if ($encoding == 'base64') {
 562         $body = base64_decode($body);
 563     }
 564
 565     // All other encodings are returned raw.
 566     return $body;
 567 }
 568
 569 /*
 570  * This functions decode strings that is encoded according to
 571  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 572  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 573  */
 574 function decodeHeader ($string, $utfencode=true,$htmlsave=true) {
 575     global $languages, $squirrelmail_language;
 576     if (is_array($string)) {
 577         $string = implode("\n", $string);
 578     }
 579
 580     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 581         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 582         $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
 583         // Do we need to return at this point?
 584         // return $string;
 585     }
 586     $i = 0;
 587     $iLastMatch = -2;
 588     $encoded = false;
 589
 590     $aString = explode(' ',$string);
 591     $ret = '';
 592     foreach ($aString as $chunk) {
 593         if ($encoded && !$chunk) {
 594             continue;
 595         } elseif (!$chunk) {
 596             $ret .= ' ';
 597             continue;
 598         }
 599         $encoded = false;
 600         /* if encoded words are not separated by a linear-space-white we still catch them */
 601         $j = $i-1;
 602         if ($chunk{0} === '=') { /* performance, saves an unnessecarry preg call */
 603         while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 604             /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
 605             if ($iLastMatch !== $j) {
 606                 if ($htmlsave) {
 607                     $ret .= '&nbsp;';
 608                 } else {
 609                     $ret .= ' ';
 610                 }
 611             }
 612             $iLastMatch = $i;
 613             $j = $i;
 614             $ret .= $res[1];
 615             $encoding = ucfirst($res[3]);
 616             switch ($encoding)
 617             {
 618             case 'B':
 619                 $replace = base64_decode($res[4]);
 620                 $ret .= charset_decode($res[2],$replace);
 621                 break;
 622             case 'Q':
 623                 $replace = str_replace('_', ' ', $res[4]);
 624                 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 625                                     $replace);
 626                 /* Only encode into entities by default. Some places
 627                  * don't need the encoding, like the compose form.
 628                  */
 629                 if ($utfencode) {
 630                     $replace = charset_decode($res[2], $replace);
 631                 } else {
 632                     if ($htmlsave) {
 633                         $replace = htmlspecialchars($replace);
 634                     }
 635                 }
 636                 $ret .= $replace;
 637                 break;
 638             default:
 639                 break;
 640             }
 641             $chunk = $res[5];
 642             $encoded = true;
 643         }
 644         }
 645         if (!$encoded) {
 646             if ($htmlsave) {
 647                 $ret .= '&nbsp;';
 648             } else {
 649                 $ret .= ' ';
 650             }
 651         }
 652
 653         if (!$encoded && $htmlsave) {
 654             $ret .= htmlspecialchars($chunk);
 655         } else {
 656             $ret .= $chunk;
 657         }
 658         ++$i;
 659     }
 660     /* remove the first added space */
 661     if ($ret) {
 662         if ($htmlsave) {
 663             $ret = substr($ret,6);
 664         } else {
 665             $ret = substr($ret,1);
 666         }
 667     }
 668
 669     return $ret;
 670 }
 671
 672 /*
 673  * Encode a string according to RFC 1522 for use in headers if it
 674  * contains 8-bit characters or anything that looks like it should
 675  * be encoded.
 676  */
 677 function encodeHeader ($string) {
 678     global $default_charset, $languages, $squirrelmail_language;
 679
 680     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 681         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 682         return  $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
 683     }
 684     if (strtolower($default_charset) == 'iso-8859-1') {
 685         $string = str_replace("\240",' ',$string);
 686     }
 687
 688     // Encode only if the string contains 8-bit characters or =?
 689     $j = strlen($string);
 690     $max_l = 75 - strlen($default_charset) - 7;
 691     $aRet = array();
 692     $ret = '';
 693     $iEncStart = $enc_init = false;
 694     $cur_l = $iOffset = 0;
 695     for($i = 0; $i < $j; ++$i) {
 696         switch($string{$i})
 697         {
 698         case '=':
 699         case '<':
 700         case '>':
 701         case ',':
 702         case '?':
 703         case '_':
 704             if ($iEncStart === false) {
 705                 $iEncStart = $i;
 706             }
 707             $cur_l+=3;
 708             if ($cur_l > ($max_l-2)) {
 709                 /* if there is an stringpart that doesn't need encoding, add it */
 710                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 711                 $aRet[] = "=?$default_charset?Q?$ret?=";
 712                 $iOffset = $i;
 713                 $cur_l = 0;
 714                 $ret = '';
 715                 $iEncStart = false;
 716             } else {
 717                 $ret .= sprintf("=%02X",ord($string{$i}));
 718             }
 719             break;
 720         case '(':
 721         case ')':
 722             if ($iEncStart !== false) {
 723                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 724                 $aRet[] = "=?$default_charset?Q?$ret?=";
 725                 $iOffset = $i;
 726                 $cur_l = 0;
 727                 $ret = '';
 728                 $iEncStart = false;
 729             }
 730             break;
 731         case ' ':
 732             if ($iEncStart !== false) {
 733                 $cur_l++;
 734                 if ($cur_l > $max_l) {
 735                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 736                     $aRet[] = "=?$default_charset?Q?$ret?=";
 737                     $iOffset = $i;
 738                     $cur_l = 0;
 739                     $ret = '';
 740                     $iEncStart = false;
 741                 } else {
 742                     $ret .= '_';
 743                 }
 744             }
 745             break;
 746         default:
 747             $k = ord($string{$i});
 748             if ($k > 126) {
 749                 if ($iEncStart === false) {
 750                     $iEncStart = $i;
 751                 }
 752                 $cur_l += 3;
 753                 /* first we add the encoded string that reached it's max size */
 754                 if ($cur_l > ($max_l-2)) {
 755                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 756                     $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
 757                     $cur_l = 3;
 758                     $ret = '';
 759                     $iOffset = $i;
 760                     $iEncStart = $i;
 761                 }
 762                 $enc_init = true;
 763                 $ret .= sprintf("=%02X", $k);
 764             } else {
 765                 if ($iEncStart !== false) {
 766                     $cur_l++;
 767                     if ($cur_l > $max_l) {
 768                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 769                         $aRet[] = "=?$default_charset?Q?$ret?=";
 770                         $iEncStart = false;
 771                         $iOffset = $i;
 772                         $cur_l = 0;
 773                         $ret = '';
 774                     } else {
 775                         $ret .= $string{$i};
 776                     }
 777                 }
 778             }
 779             break;
 780         }
 781     }
 782
 783     if ($enc_init) {
 784         if ($iEncStart !== false) {
 785             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 786             $aRet[] = "=?$default_charset?Q?$ret?=";
 787         } else {
 788             $aRet[] = substr($string,$iOffset);
 789         }
 790         $string = implode('',$aRet);
 791     }
 792     return $string;
 793 }
 794
 795 /* This function trys to locate the entity_id of a specific mime element */
 796 function find_ent_id($id, $message) {
 797     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
 798         if ($message->entities[$i]->header->type0 == 'multipart')  {
 799             $ret = find_ent_id($id, $message->entities[$i]);
 800         } else {
 801             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
 802 //                if (sq_check_save_extension($message->entities[$i])) {
 803                     return $message->entities[$i]->entity_id;
 804 //                }
 805             }
 806         }
 807     }
 808     return $ret;
 809 }
 810
 811 function sq_check_save_extension($message) {
 812     $filename = $message->getFilename();
 813     $ext = substr($filename, strrpos($filename,'.')+1);
 814     $save_extensions = array('jpg','jpeg','gif','png','bmp');
 815     return in_array($ext, $save_extensions);
 816 }
 817
 818
 819 /**
 820  ** HTMLFILTER ROUTINES
 821  */
 822
 823 /**
 824  * This function returns the final tag out of the tag name, an array
 825  * of attributes, and the type of the tag. This function is called by
 826  * sq_sanitize internally.
 827  *
 828  * @param  $tagname  the name of the tag.
 829  * @param  $attary   the array of attributes and their values
 830  * @param  $tagtype  The type of the tag (see in comments).
 831  * @return           a string with the final tag representation.
 832  */
 833 function sq_tagprint($tagname, $attary, $tagtype){
 834     $me = 'sq_tagprint';
 835
 836     if ($tagtype == 2){
 837         $fulltag = '</' . $tagname . '>';
 838     } else {
 839         $fulltag = '<' . $tagname;
 840         if (is_array($attary) && sizeof($attary)){
 841             $atts = Array();
 842             while (list($attname, $attvalue) = each($attary)){
 843                 array_push($atts, "$attname=$attvalue");
 844             }
 845             $fulltag .= ' ' . join(" ", $atts);
 846         }
 847         if ($tagtype == 3){
 848             $fulltag .= ' /';
 849         }
 850         $fulltag .= '>';
 851     }
 852     return $fulltag;
 853 }
 854
 855 /**
 856  * A small helper function to use with array_walk. Modifies a by-ref
 857  * value and makes it lowercase.
 858  *
 859  * @param  $val a value passed by-ref.
 860  * @return      void since it modifies a by-ref value.
 861  */
 862 function sq_casenormalize(&$val){
 863     $val = strtolower($val);
 864 }
 865
 866 /**
 867  * This function skips any whitespace from the current position within
 868  * a string and to the next non-whitespace value.
 869  *
 870  * @param  $body   the string
 871  * @param  $offset the offset within the string where we should start
 872  *                 looking for the next non-whitespace character.
 873  * @return         the location within the $body where the next
 874  *                 non-whitespace char is located.
 875  */
 876 function sq_skipspace($body, $offset){
 877     $me = 'sq_skipspace';
 878     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
 879     if (sizeof($matches{1})){
 880         $count = strlen($matches{1});
 881         $offset += $count;
 882     }
 883     return $offset;
 884 }
 885
 886 /**
 887  * This function looks for the next character within a string.  It's
 888  * really just a glorified "strpos", except it catches if failures
 889  * nicely.
 890  *
 891  * @param  $body   The string to look for needle in.
 892  * @param  $offset Start looking from this position.
 893  * @param  $needle The character/string to look for.
 894  * @return         location of the next occurance of the needle, or
 895  *                 strlen($body) if needle wasn't found.
 896  */
 897 function sq_findnxstr($body, $offset, $needle){
 898     $me  = 'sq_findnxstr';
 899     $pos = strpos($body, $needle, $offset);
 900     if ($pos === FALSE){
 901         $pos = strlen($body);
 902     }
 903     return $pos;
 904 }
 905
 906 /**
 907  * This function takes a PCRE-style regexp and tries to match it
 908  * within the string.
 909  *
 910  * @param  $body   The string to look for needle in.
 911  * @param  $offset Start looking from here.
 912  * @param  $reg    A PCRE-style regex to match.
 913  * @return         Returns a false if no matches found, or an array
 914  *                 with the following members:
 915  *                 - integer with the location of the match within $body
 916  *                 - string with whatever content between offset and the match
 917  *                 - string with whatever it is we matched
 918  */
 919 function sq_findnxreg($body, $offset, $reg){
 920     $me = 'sq_findnxreg';
 921     $matches = Array();
 922     $retarr = Array();
 923     preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
 924     if (!$matches{0}){
 925         $retarr = false;
 926     } else {
 927         $retarr{0} = $offset + strlen($matches{1});
 928         $retarr{1} = $matches{1};
 929         $retarr{2} = $matches{2};
 930     }
 931     return $retarr;
 932 }
 933
 934 /**
 935  * This function looks for the next tag.
 936  *
 937  * @param  $body   String where to look for the next tag.
 938  * @param  $offset Start looking from here.
 939  * @return         false if no more tags exist in the body, or
 940  *                 an array with the following members:
 941  *                 - string with the name of the tag
 942  *                 - array with attributes and their values
 943  *                 - integer with tag type (1, 2, or 3)
 944  *                 - integer where the tag starts (starting "<")
 945  *                 - integer where the tag ends (ending ">")
 946  *                 first three members will be false, if the tag is invalid.
 947  */
 948 function sq_getnxtag($body, $offset){
 949     $me = 'sq_getnxtag';
 950     if ($offset > strlen($body)){
 951         return false;
 952     }
 953     $lt = sq_findnxstr($body, $offset, "<");
 954     if ($lt == strlen($body)){
 955         return false;
 956     }
 957     /**
 958      * We are here:
 959      * blah blah <tag attribute="value">
 960      * \---------^
 961      */
 962     $pos = sq_skipspace($body, $lt+1);
 963     if ($pos >= strlen($body)){
 964         return Array(false, false, false, $lt, strlen($body));
 965     }
 966     /**
 967      * There are 3 kinds of tags:
 968      * 1. Opening tag, e.g.:
 969      *    <a href="blah">
 970      * 2. Closing tag, e.g.:
 971      *    </a>
 972      * 3. XHTML-style content-less tag, e.g.:
 973      *    <img src="blah"/>
 974      */
 975     $tagtype = false;
 976     switch (substr($body, $pos, 1)){
 977         case '/':
 978             $tagtype = 2;
 979             $pos++;
 980             break;
 981         case '!':
 982             /**
 983              * A comment or an SGML declaration.
 984              */
 985             if (substr($body, $pos+1, 2) == "--"){
 986                 $gt = strpos($body, "-->", $pos);
 987                 if ($gt === false){
 988                     $gt = strlen($body);
 989                 } else {
 990                     $gt += 2;
 991                 }
 992                 return Array(false, false, false, $lt, $gt);
 993             } else {
 994                 $gt = sq_findnxstr($body, $pos, ">");
 995                 return Array(false, false, false, $lt, $gt);
 996             }
 997             break;
 998         default:
 999             /**
1000              * Assume tagtype 1 for now. If it's type 3, we'll switch values
1001              * later.
1002              */
1003             $tagtype = 1;
1004             break;
1005     }
1006
1007     $tag_start = $pos;
1008     $tagname = '';
1009     /**
1010      * Look for next [\W-_], which will indicate the end of the tag name.
1011      */
1012     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1013     if ($regary == false){
1014         return Array(false, false, false, $lt, strlen($body));
1015     }
1016     list($pos, $tagname, $match) = $regary;
1017     $tagname = strtolower($tagname);
1018
1019     /**
1020      * $match can be either of these:
1021      * '>'  indicating the end of the tag entirely.
1022      * '\s' indicating the end of the tag name.
1023      * '/'  indicating that this is type-3 xhtml tag.
1024      *
1025      * Whatever else we find there indicates an invalid tag.
1026      */
1027     switch ($match){
1028         case '/':
1029             /**
1030              * This is an xhtml-style tag with a closing / at the
1031              * end, like so: <img src="blah"/>. Check if it's followed
1032              * by the closing bracket. If not, then this tag is invalid
1033              */
1034             if (substr($body, $pos, 2) == "/>"){
1035                 $pos++;
1036                 $tagtype = 3;
1037             } else {
1038                 $gt = sq_findnxstr($body, $pos, ">");
1039                 $retary = Array(false, false, false, $lt, $gt);
1040                 return $retary;
1041             }
1042         case '>':
1043             return Array($tagname, false, $tagtype, $lt, $pos);
1044             break;
1045         default:
1046             /**
1047              * Check if it's whitespace
1048              */
1049             if (!preg_match('/\s/', $match)){
1050                 /**
1051                  * This is an invalid tag! Look for the next closing ">".
1052                  */
1053                 $gt = sq_findnxstr($body, $offset, ">");
1054                 return Array(false, false, false, $lt, $gt);
1055             }
1056             break;
1057     }
1058
1059     /**
1060      * At this point we're here:
1061      * <tagname  attribute='blah'>
1062      * \-------^
1063      *
1064      * At this point we loop in order to find all attributes.
1065      */
1066     $attname = '';
1067     $atttype = false;
1068     $attary = Array();
1069
1070     while ($pos <= strlen($body)){
1071         $pos = sq_skipspace($body, $pos);
1072         if ($pos == strlen($body)){
1073             /**
1074              * Non-closed tag.
1075              */
1076             return Array(false, false, false, $lt, $pos);
1077         }
1078         /**
1079          * See if we arrived at a ">" or "/>", which means that we reached
1080          * the end of the tag.
1081          */
1082         $matches = Array();
1083         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1084             /**
1085              * Yep. So we did.
1086              */
1087             $pos += strlen($matches{1});
1088             if ($matches{2} == "/>"){
1089                 $tagtype = 3;
1090                 $pos++;
1091             }
1092             return Array($tagname, $attary, $tagtype, $lt, $pos);
1093         }
1094
1095         /**
1096          * There are several types of attributes, with optional
1097          * [:space:] between members.
1098          * Type 1:
1099          *   attrname[:space:]=[:space:]'CDATA'
1100          * Type 2:
1101          *   attrname[:space:]=[:space:]"CDATA"
1102          * Type 3:
1103          *   attr[:space:]=[:space:]CDATA
1104          * Type 4:
1105          *   attrname
1106          *
1107          * We leave types 1 and 2 the same, type 3 we check for
1108          * '"' and convert to "&quot" if needed, then wrap in
1109          * double quotes. Type 4 we convert into:
1110          * attrname="yes".
1111          */
1112         $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1113         if ($regary == false){
1114             /**
1115              * Looks like body ended before the end of tag.
1116              */
1117             return Array(false, false, false, $lt, strlen($body));
1118         }
1119         list($pos, $attname, $match) = $regary;
1120         $attname = strtolower($attname);
1121         /**
1122          * We arrived at the end of attribute name. Several things possible
1123          * here:
1124          * '>'  means the end of the tag and this is attribute type 4
1125          * '/'  if followed by '>' means the same thing as above
1126          * '\s' means a lot of things -- look what it's followed by.
1127          *      anything else means the attribute is invalid.
1128          */
1129         switch($match){
1130             case '/':
1131                 /**
1132                  * This is an xhtml-style tag with a closing / at the
1133                  * end, like so: <img src="blah"/>. Check if it's followed
1134                  * by the closing bracket. If not, then this tag is invalid
1135                  */
1136                 if (substr($body, $pos, 2) == "/>"){
1137                     $pos++;
1138                     $tagtype = 3;
1139                 } else {
1140                     $gt = sq_findnxstr($body, $pos, ">");
1141                     $retary = Array(false, false, false, $lt, $gt);
1142                     return $retary;
1143                 }
1144             case '>':
1145                 $attary{$attname} = '"yes"';
1146                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1147                 break;
1148             default:
1149                 /**
1150                  * Skip whitespace and see what we arrive at.
1151                  */
1152                 $pos = sq_skipspace($body, $pos);
1153                 $char = substr($body, $pos, 1);
1154                 /**
1155                  * Two things are valid here:
1156                  * '=' means this is attribute type 1 2 or 3.
1157                  * \w means this was attribute type 4.
1158                  * anything else we ignore and re-loop. End of tag and
1159                  * invalid stuff will be caught by our checks at the beginning
1160                  * of the loop.
1161                  */
1162                 if ($char == "="){
1163                     $pos++;
1164                     $pos = sq_skipspace($body, $pos);
1165                     /**
1166                      * Here are 3 possibilities:
1167                      * "'"  attribute type 1
1168                      * '"'  attribute type 2
1169                      * everything else is the content of tag type 3
1170                      */
1171                     $quot = substr($body, $pos, 1);
1172                     if ($quot == "'"){
1173                         $regary = sq_findnxreg($body, $pos+1, "\'");
1174                         if ($regary == false){
1175                             return Array(false, false, false, $lt, strlen($body));
1176                         }
1177                         list($pos, $attval, $match) = $regary;
1178                         $pos++;
1179                         $attary{$attname} = "'" . $attval . "'";
1180                     } else if ($quot == '"'){
1181                         $regary = sq_findnxreg($body, $pos+1, '\"');
1182                         if ($regary == false){
1183                             return Array(false, false, false, $lt, strlen($body));
1184                         }
1185                         list($pos, $attval, $match) = $regary;
1186                         $pos++;
1187                         $attary{$attname} = '"' . $attval . '"';
1188                     } else {
1189                         /**
1190                          * These are hateful. Look for \s, or >.
1191                          */
1192                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1193                         if ($regary == false){
1194                             return Array(false, false, false, $lt, strlen($body));
1195                         }
1196                         list($pos, $attval, $match) = $regary;
1197                         /**
1198                          * If it's ">" it will be caught at the top.
1199                          */
1200                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1201                         $attary{$attname} = '"' . $attval . '"';
1202                     }
1203                 } else if (preg_match("|[\w/>]|", $char)) {
1204                     /**
1205                      * That was attribute type 4.
1206                      */
1207                     $attary{$attname} = '"yes"';
1208                 } else {
1209                     /**
1210                      * An illegal character. Find next '>' and return.
1211                      */
1212                     $gt = sq_findnxstr($body, $pos, ">");
1213                     return Array(false, false, false, $lt, $gt);
1214                 }
1215                 break;
1216         }
1217     }
1218     /**
1219      * The fact that we got here indicates that the tag end was never
1220      * found. Return invalid tag indication so it gets stripped.
1221      */
1222     return Array(false, false, false, $lt, strlen($body));
1223 }
1224
1225 /**
1226  * This function checks attribute values for entity-encoded values
1227  * and returns them translated into 8-bit strings so we can run
1228  * checks on them.
1229  *
1230  * @param  $attvalue A string to run entity check against.
1231  * @return           Translated value.
1232  */
1233 function sq_deent($attvalue){
1234     $me = 'sq_deent';
1235     /**
1236      * See if we have to run the checks first. All entities must start
1237      * with "&".
1238      */
1239     if (strpos($attvalue, "&") === false){
1240         return $attvalue;
1241     }
1242     /**
1243      * Check named entities first.
1244      */
1245     $trans = get_html_translation_table(HTML_ENTITIES);
1246     /**
1247      * Leave &quot; in, as it can mess us up.
1248      */
1249     $trans = array_flip($trans);
1250     unset($trans{"&quot;"});
1251     while (list($ent, $val) = each($trans)){
1252         $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1253     }
1254     /**
1255      * Now translate numbered entities from 1 to 255 if needed.
1256      */
1257     if (strpos($attvalue, "#") !== false){
1258         $omit = Array(34, 39);
1259         for ($asc=1; $asc<256; $asc++){
1260             if (!in_array($asc, $omit)){
1261                 $chr = chr($asc);
1262                 $attvalue = preg_replace("/\&#0*$asc;*(\D)/si", "$chr\\1",
1263                                          $attvalue);
1264                 $attvalue = preg_replace("/\&#x0*".dechex($asc).";*(\W)/si",
1265                                          "$chr\\1", $attvalue);
1266             }
1267         }
1268     }
1269     return $attvalue;
1270 }
1271
1272 /**
1273  * This function runs various checks against the attributes.
1274  *
1275  * @param  $tagname         String with the name of the tag.
1276  * @param  $attary          Array with all tag attributes.
1277  * @param  $rm_attnames     See description for sq_sanitize
1278  * @param  $bad_attvals     See description for sq_sanitize
1279  * @param  $add_attr_to_tag See description for sq_sanitize
1280  * @param  $message         message object
1281  * @param  $id              message id
1282  * @return                  Array with modified attributes.
1283  */
1284 function sq_fixatts($tagname,
1285                     $attary,
1286                     $rm_attnames,
1287                     $bad_attvals,
1288                     $add_attr_to_tag,
1289                     $message,
1290                     $id,
1291                     $mailbox
1292                     ){
1293     $me = 'sq_fixatts';
1294     while (list($attname, $attvalue) = each($attary)){
1295         /**
1296          * See if this attribute should be removed.
1297          */
1298         foreach ($rm_attnames as $matchtag=>$matchattrs){
1299             if (preg_match($matchtag, $tagname)){
1300                 foreach ($matchattrs as $matchattr){
1301                     if (preg_match($matchattr, $attname)){
1302                         unset($attary{$attname});
1303                         continue;
1304                     }
1305                 }
1306             }
1307         }
1308         /**
1309          * Remove any entities.
1310          */
1311         $attvalue = sq_deent($attvalue);
1312
1313         /**
1314          * Now let's run checks on the attvalues.
1315          * I don't expect anyone to comprehend this. If you do,
1316          * get in touch with me so I can drive to where you live and
1317          * shake your hand personally. :)
1318          */
1319         foreach ($bad_attvals as $matchtag=>$matchattrs){
1320             if (preg_match($matchtag, $tagname)){
1321                 foreach ($matchattrs as $matchattr=>$valary){
1322                     if (preg_match($matchattr, $attname)){
1323                         /**
1324                          * There are two arrays in valary.
1325                          * First is matches.
1326                          * Second one is replacements
1327                          */
1328                         list($valmatch, $valrepl) = $valary;
1329                         $newvalue =
1330                             preg_replace($valmatch, $valrepl, $attvalue);
1331                         if ($newvalue != $attvalue){
1332                             $attary{$attname} = $newvalue;
1333                         }
1334                     }
1335                 }
1336             }
1337         }
1338         /**
1339          * Turn cid: urls into http-friendly ones.
1340          */
1341         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1342             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1343         }
1344     }
1345     /**
1346      * See if we need to append any attributes to this tag.
1347      */
1348     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1349         if (preg_match($matchtag, $tagname)){
1350             $attary = array_merge($attary, $addattary);
1351         }
1352     }
1353     return $attary;
1354 }
1355
1356 /**
1357  * This function edits the style definition to make them friendly and
1358  * usable in squirrelmail.
1359  *
1360  * @param  $message  the message object
1361  * @param  $id       the message id
1362  * @param  $content  a string with whatever is between <style> and </style>
1363  * @return           a string with edited content.
1364  */
1365 function sq_fixstyle($message, $id, $content){
1366     global $view_unsafe_images;
1367     $me = 'sq_fixstyle';
1368     /**
1369      * First look for general BODY style declaration, which would be
1370      * like so:
1371      * body {background: blah-blah}
1372      * and change it to .bodyclass so we can just assign it to a <div>
1373      */
1374     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1375     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1376     /**
1377      * Fix url('blah') declarations.
1378      */
1379     $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1380                             "url(\\1$secremoveimg\\2)", $content);
1381     /**
1382      * Fix url('https*://.*) declarations but only if $view_unsafe_images
1383      * is false.
1384      */
1385     if (!$view_unsafe_images){
1386         $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1387                                 "url(\\1$secremoveimg\\2)", $content);
1388     }
1389
1390     /**
1391      * Fix urls that refer to cid:
1392      */
1393     while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1394                       $matches)){
1395         $cidurl = $matches{1};
1396         $httpurl = sq_cid2http($message, $id, $cidurl);
1397         $content = preg_replace("|url\($cidurl\)|si",
1398                                 "url($httpurl)", $content);
1399     }
1400
1401     /**
1402      * Fix stupid css declarations which lead to vulnerabilities
1403      * in IE.
1404      */
1405     $match   = Array('/expression/i',
1406                      '/behaviou*r/i',
1407                      '/binding/i');
1408     $replace = Array('idiocy', 'idiocy', 'idiocy');
1409     $content = preg_replace($match, $replace, $content);
1410     return $content;
1411 }
1412
1413 /**
1414  * This function converts cid: url's into the ones that can be viewed in
1415  * the browser.
1416  *
1417  * @param  $message  the message object
1418  * @param  $id       the message id
1419  * @param  $cidurl   the cid: url.
1420  * @return           a string with a http-friendly url
1421  */
1422 function sq_cid2http($message, $id, $cidurl, $mailbox){
1423     /**
1424      * Get rid of quotes.
1425      */
1426     $quotchar = substr($cidurl, 0, 1);
1427     $cidurl = str_replace($quotchar, "", $cidurl);
1428     $cidurl = substr(trim($cidurl), 4);
1429     $linkurl = find_ent_id($cidurl, $message);
1430     /* in case of non-save cid links $httpurl should be replaced by a sort of
1431        unsave link image */
1432     $httpurl = '';
1433     if ($linkurl) {
1434         $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1435                    "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1436                    '&amp;ent_id=' . $linkurl . $quotchar;
1437     }
1438     return $httpurl;
1439 }
1440
1441 /**
1442  * This function changes the <body> tag into a <div> tag since we
1443  * can't really have a body-within-body.
1444  *
1445  * @param  $attary  an array of attributes and values of <body>
1446  * @return          a modified array of attributes to be set for <div>
1447  */
1448 function sq_body2div($attary){
1449     $me = 'sq_body2div';
1450     $divattary = Array('class' => "'bodyclass'");
1451     $bgcolor = '#ffffff';
1452     $text = '#000000';
1453     $styledef = '';
1454     if (is_array($attary) && sizeof($attary) > 0){
1455         foreach ($attary as $attname=>$attvalue){
1456             $quotchar = substr($attvalue, 0, 1);
1457             $attvalue = str_replace($quotchar, "", $attvalue);
1458             switch ($attname){
1459                 case 'background':
1460                     $styledef .= "background-image: url('$attvalue'); ";
1461                     break;
1462                 case 'bgcolor':
1463                     $styledef .= "background-color: $attvalue; ";
1464                     break;
1465                 case 'text':
1466                     $styledef .= "color: $attvalue; ";
1467                     break;
1468             }
1469         }
1470         if (strlen($styledef) > 0){
1471             $divattary{"style"} = "\"$styledef\"";
1472         }
1473     }
1474     return $divattary;
1475 }
1476
1477 /**
1478  * This is the main function and the one you should actually be calling.
1479  * There are several variables you should be aware of an which need
1480  * special description.
1481  *
1482  * Since the description is quite lengthy, see it here:
1483  * http://www.mricon.com/html/phpfilter.html
1484  *
1485  * @param $body                 the string with HTML you wish to filter
1486  * @param $tag_list             see description above
1487  * @param $rm_tags_with_content see description above
1488  * @param $self_closing_tags    see description above
1489  * @param $force_tag_closing    see description above
1490  * @param $rm_attnames          see description above
1491  * @param $bad_attvals          see description above
1492  * @param $add_attr_to_tag      see description above
1493  * @param $message              message object
1494  * @param $id                   message id
1495  * @return                      sanitized html safe to show on your pages.
1496  */
1497 function sq_sanitize($body,
1498                      $tag_list,
1499                      $rm_tags_with_content,
1500                      $self_closing_tags,
1501                      $force_tag_closing,
1502                      $rm_attnames,
1503                      $bad_attvals,
1504                      $add_attr_to_tag,
1505                      $message,
1506                      $id,
1507                      $mailbox
1508                      ){
1509     $me = 'sq_sanitize';
1510     /**
1511      * Normalize rm_tags and rm_tags_with_content.
1512      */
1513     @array_walk($rm_tags, 'sq_casenormalize');
1514     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1515     @array_walk($self_closing_tags, 'sq_casenormalize');
1516     /**
1517      * See if tag_list is of tags to remove or tags to allow.
1518      * false  means remove these tags
1519      * true   means allow these tags
1520      */
1521     $rm_tags = array_shift($tag_list);
1522     $curpos = 0;
1523     $open_tags = Array();
1524     $trusted = "<!-- begin sanitized html -->\n";
1525     $skip_content = false;
1526     /**
1527      * Take care of netscape's stupid javascript entities like
1528      * &{alert('boo')};
1529      */
1530     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1531
1532     while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1533         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1534         $free_content = substr($body, $curpos, $lt-$curpos);
1535         /**
1536          * Take care of <style>
1537          */
1538         if ($tagname == "style" && $tagtype == 2){
1539             /**
1540              * This is a closing </style>. Edit the
1541              * content before we apply it.
1542              */
1543             $free_content = sq_fixstyle($message, $id, $free_content);
1544         }
1545         if ($skip_content == false){
1546             $trusted .= $free_content;
1547         }
1548         if ($tagname != FALSE){
1549             if ($tagtype == 2){
1550                 if ($skip_content == $tagname){
1551                     /**
1552                      * Got to the end of tag we needed to remove.
1553                      */
1554                     $tagname = false;
1555                     $skip_content = false;
1556                 } else {
1557                     if ($skip_content == false){
1558                         if ($tagname == "body"){
1559                             $tagname = "div";
1560                         } else {
1561                             if (isset($open_tags{$tagname}) &&
1562                                 $open_tags{$tagname} > 0){
1563                                 $open_tags{$tagname}--;
1564                             } else {
1565                                 $tagname = false;
1566                             }
1567                         }
1568                     }
1569                 }
1570             } else {
1571                 /**
1572                  * $rm_tags_with_content
1573                  */
1574                 if ($skip_content == false){
1575                     /**
1576                      * See if this is a self-closing type and change
1577                      * tagtype appropriately.
1578                      */
1579                     if ($tagtype == 1
1580                         && in_array($tagname, $self_closing_tags)){
1581                         $tagtype=3;
1582                     }
1583                     /**
1584                      * See if we should skip this tag and any content
1585                      * inside it.
1586                      */
1587                     if ($tagtype == 1 &&
1588                         in_array($tagname, $rm_tags_with_content)){
1589                         $skip_content = $tagname;
1590                     } else {
1591                         if (($rm_tags == false
1592                              && in_array($tagname, $tag_list)) ||
1593                             ($rm_tags == true &&
1594                              !in_array($tagname, $tag_list))){
1595                             $tagname = false;
1596                         } else {
1597                             if ($tagtype == 1){
1598                                 if (isset($open_tags{$tagname})){
1599                                     $open_tags{$tagname}++;
1600                                 } else {
1601                                     $open_tags{$tagname}=1;
1602                                 }
1603                             }
1604                             /**
1605                              * This is where we run other checks.
1606                              */
1607                             if (is_array($attary) && sizeof($attary) > 0){
1608                                 $attary = sq_fixatts($tagname,
1609                                                      $attary,
1610                                                      $rm_attnames,
1611                                                      $bad_attvals,
1612                                                      $add_attr_to_tag,
1613                                                      $message,
1614                                                      $id,
1615                                                      $mailbox
1616                                                      );
1617                             }
1618                             /**
1619                              * Convert body into div.
1620                              */
1621                             if ($tagname == "body"){
1622                                 $tagname = "div";
1623                                 $attary = sq_body2div($attary, $message, $id);
1624                             }
1625                         }
1626                     }
1627                 }
1628             }
1629             if ($tagname != false && $skip_content == false){
1630                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1631             }
1632         }
1633         $curpos = $gt+1;
1634     }
1635     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1636     if ($force_tag_closing == true){
1637         foreach ($open_tags as $tagname=>$opentimes){
1638             while ($opentimes > 0){
1639                 $trusted .= '</' . $tagname . '>';
1640                 $opentimes--;
1641             }
1642         }
1643         $trusted .= "\n";
1644     }
1645     $trusted .= "<!-- end sanitized html -->\n";
1646     return $trusted;
1647 }
1648
1649 /**
1650  * This is a wrapper function to call html sanitizing routines.
1651  *
1652  * @param  $body  the body of the message
1653  * @param  $id    the id of the message
1654  * @return        a string with html safe to display in the browser.
1655  */
1656 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1657     global $attachment_common_show_images, $view_unsafe_images,
1658            $has_unsafe_images;
1659     /**
1660      * Don't display attached images in HTML mode.
1661      */
1662     $attachment_common_show_images = false;
1663     $tag_list = Array(
1664                       false,
1665                       "object",
1666                       "meta",
1667                       "html",
1668                       "head",
1669                       "base",
1670                       "link",
1671                       "frame",
1672                       "iframe"
1673                       );
1674
1675     $rm_tags_with_content = Array(
1676                                   "script",
1677                                   "applet",
1678                                   "embed",
1679                                   "title"
1680                                   );
1681
1682     $self_closing_tags =  Array(
1683                                 "img",
1684                                 "br",
1685                                 "hr",
1686                                 "input"
1687                                 );
1688
1689     $force_tag_closing = false;
1690
1691     $rm_attnames = Array(
1692                          "/.*/" =>
1693                          Array(
1694                                "/target/i",
1695                                "/^on.*/i",
1696                                "/^dynsrc/i",
1697                                "/^data.*/i",
1698                                "/^lowsrc.*/i"
1699                                )
1700                          );
1701
1702     $secremoveimg = "../images/" . _("sec_remove_eng.png");
1703     $bad_attvals = Array(
1704         "/.*/" =>
1705             Array(
1706                 "/^src|background/i" =>
1707                     Array(
1708                           Array(
1709                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1710                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1711                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1712                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1713                                 ),
1714                           Array(
1715                                 "\\1$secremoveimg\\2",
1716                                 "\\1$secremoveimg\\2",
1717                                 "\\1$secremoveimg\\2",
1718                                 "\\1$secremoveimg\\2"
1719                                 )
1720                         ),
1721                 "/^href|action/i" =>
1722                     Array(
1723                           Array(
1724                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1725                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1726                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1727                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1728                                 ),
1729                           Array(
1730                                 "\\1#\\2",
1731                                 "\\1#\\2",
1732                                 "\\1#\\2",
1733                                 "\\1#\\2"
1734                                 )
1735                         ),
1736                 "/^style/i" =>
1737                     Array(
1738                           Array(
1739                                 "/expression/i",
1740                                 "/binding/i",
1741                                 "/behaviou*r/i",
1742                                 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
1743                                 "/url\(([\'\"])\s*\S+script\s*:.*([\'\"])\)/si",
1744                                 "/url\(([\'\"])\s*mocha\s*:.*([\'\"])\)/si",
1745                                 "/url\(([\'\"])\s*about\s*:.*([\'\"])\)/si"
1746                                ),
1747                           Array(
1748                                 "idiocy",
1749                                 "idiocy",
1750                                 "idiocy",
1751                                 "url(\\1#\\2)",
1752                                 "url(\\1#\\2)",
1753                                 "url(\\1#\\2)",
1754                                 "url(\\1#\\2)"
1755                                )
1756                           )
1757                 )
1758         );
1759     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1760         $view_unsafe_images = false;
1761     }
1762     if (!$view_unsafe_images){
1763         /**
1764          * Remove any references to http/https if view_unsafe_images set
1765          * to false.
1766          */
1767          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1768                     '/^([\'\"])\s*https*:.*([\'\"])/si');
1769          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1770                     "\\1$secremoveimg\\2");
1771          array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
1772                     '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1773          array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
1774                     "url(\\1$secremoveimg\\2)");
1775     }
1776
1777     $add_attr_to_tag = Array(
1778                              "/^a$/i" => Array('target'=>'"_new"')
1779                              );
1780     $trusted = sq_sanitize($body,
1781                            $tag_list,
1782                            $rm_tags_with_content,
1783                            $self_closing_tags,
1784                            $force_tag_closing,
1785                            $rm_attnames,
1786                            $bad_attvals,
1787                            $add_attr_to_tag,
1788                            $message,
1789                            $id,
1790                            $mailbox
1791                            );
1792     if (preg_match("|$secremoveimg|i", $trusted)){
1793         $has_unsafe_images = true;
1794     }
1795     return $trusted;
1796 }
1797
1798 ?>