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         $errormessage  = _("SquirrelMail could not decode the bodystructure of the message");
  40         $errormessage .= '<BR>'._("the provided bodystructure by your imap-server").':<BR><BR>';
  41         $errormessage .= '<pre>' . htmlspecialchars($read) . '</pre>';
  42         plain_error_message( $errormessage, $color );
  43         echo '</body></html>';
  44         exit;
  45     }
  46     if (count($flags)) {
  47         foreach ($flags as $flag) {
  48             $char = strtoupper($flag{1});
  49             switch ($char) {
  50                 case 'S':
  51                     if (strtolower($flag) == '\\seen') {
  52                         $msg->is_seen = true;
  53                     }
  54                     break;
  55                 case 'A':
  56                     if (strtolower($flag) == '\\answered') {
  57                         $msg->is_answered = true;
  58                     }
  59                     break;
  60                 case 'D':
  61                     if (strtolower($flag) == '\\deleted') {
  62                         $msg->is_deleted = true;
  63                     }
  64                     break;
  65                 case 'F':
  66                     if (strtolower($flag) == '\\flagged') {
  67                         $msg->is_flagged = true;
  68                     }
  69                     break;
  70                 case 'M':
  71                     if (strtolower($flag) == '$mdnsent') {
  72                         $msg->is_mdnsent = true;
  73                     }
  74                     break;
  75                 default:
  76                     break;
  77             }
  78         }
  79     }
  80     //    listEntities($msg);
  81     return $msg;
  82 }
  83
  84
  85
  86 /* This starts the parsing of a particular structure.  It is called recursively,
  87  * so it can be passed different structures.  It returns an object of type
  88  * $message.
  89  * First, it checks to see if it is a multipart message.  If it is, then it
  90  * handles that as it sees is necessary.  If it is just a regular entity,
  91  * then it parses it and adds the necessary header information (by calling out
  92  * to mime_get_elements()
  93  */
  94
  95 function mime_fetch_body($imap_stream, $id, $ent_id=1) {
  96     global $uid_support;
  97     /* Do a bit of error correction.  If we couldn't find the entity id, just guess
  98      * that it is the first one.  That is usually the case anyway.
  99      */
 100
 101     if (!$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 $uid_support;
 157
 158     /* Don't kill the connection if the browser is over a dialup
 159      * and it would take over 30 seconds to download it.
 160      * Don´t call set_time_limit in safe mode.
 161      */
 162
 163     if (!ini_get('safe_mode')) {
 164         set_time_limit(0);
 165     }
 166     /* in case of base64 encoded attachments, do not buffer them.
 167        Instead, echo the decoded attachment directly to screen */
 168     if (strtolower($encoding) == 'base64') {
 169         if (!$ent_id) {
 170            $query = "FETCH $id BODY[]";
 171         } else {
 172            $query = "FETCH $id BODY[$ent_id]";
 173         }
 174         sqimap_run_command($imap_stream,$query,true,$response,$message,$uid_support,'sqimap_base64_decode','php://stdout',true);
 175     } else {
 176        $body = mime_fetch_body ($imap_stream, $id, $ent_id);
 177        echo decodeBody($body, $encoding);
 178     }
 179
 180     /*
 181        TODO, use the same method for quoted printable.
 182        However, I assume that quoted printable attachments aren't that large
 183        so the performancegain / memory usage drop will be minimal.
 184        If we decide to add that then we need to adapt sqimap_fread because
 185        we need to split te result on \n and fread doesn't stop at \n. That
 186        means we also should provide $results from sqimap_fread (by ref) to
 187        te function and set $no_return to false. The $filter function for
 188        quoted printable should handle unsetting of $results.
 189     */
 190     /*
 191        TODO 2: find out how we write to the output stream php://stdout. fwrite
 192        doesn't work because 'php://stdout isn't a stream.
 193     */
 194
 195     return;
 196 /*
 197     fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
 198     $cnt = 0;
 199     $continue = true;
 200     $read = fgets ($imap_stream,8192);
 201
 202
 203     // This could be bad -- if the section has sqimap_session_id() . ' OK'
 204     // or similar, it will kill the download.
 205     while (!ereg("^".$sid_s." (OK|BAD|NO)(.*)$", $read, $regs)) {
 206         if (trim($read) == ')==') {
 207             $read1 = $read;
 208             $read = fgets ($imap_stream,4096);
 209             if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
 210                 return;
 211             } else {
 212                 echo decodeBody($read1, $encoding) .
 213                      decodeBody($read, $encoding);
 214             }
 215         } else if ($cnt) {
 216             echo decodeBody($read, $encoding);
 217         }
 218         $read = fgets ($imap_stream,4096);
 219         $cnt++;
 220 //      break;
 221     }
 222 */
 223 }
 224
 225 /* -[ END MIME DECODING ]----------------------------------------------------------- */
 226
 227 /* This is here for debugging purposes.  It will print out a list
 228  * of all the entity IDs that are in the $message object.
 229  */
 230 function listEntities ($message) {
 231     if ($message) {
 232         echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br>';
 233         for ($i = 0; isset($message->entities[$i]); $i++) {
 234             echo "$i : ";
 235             $msg = listEntities($message->entities[$i]);
 236
 237             if ($msg) {
 238                 echo "return: ";
 239                 return $msg;
 240             }
 241         }
 242     }
 243 }
 244
 245 function getPriorityStr($priority) {
 246     $priority_level = substr($priority,0,1);
 247
 248     switch($priority_level) {
 249         /* Check for a higher then normal priority. */
 250         case '1':
 251         case '2':
 252             $priority_string = _("High");
 253             break;
 254
 255         /* Check for a lower then normal priority. */
 256         case '4':
 257         case '5':
 258             $priority_string = _("Low");
 259             break;
 260
 261         /* Check for a normal priority. */
 262         case '3':
 263         default:
 264             $priority_level = '3';
 265             $priority_string = _("Normal");
 266             break;
 267
 268     }
 269     return $priority_string;
 270 }
 271
 272 /* returns a $message object for a particular entity id */
 273 function getEntity ($message, $ent_id) {
 274     return $message->getEntity($ent_id);
 275 }
 276
 277 /* translateText
 278  * Extracted from strings.php 23/03/2002
 279  */
 280
 281 function translateText(&$body, $wrap_at, $charset) {
 282     global $where, $what;   /* from searching */
 283     global $color;          /* color theme */
 284
 285     require_once(SM_PATH . 'functions/url_parser.php');
 286
 287     $body_ary = explode("\n", $body);
 288     for ($i=0; $i < count($body_ary); $i++) {
 289         $line = $body_ary[$i];
 290         if (strlen($line) - 2 >= $wrap_at) {
 291             sqWordWrap($line, $wrap_at);
 292         }
 293         $line = charset_decode($charset, $line);
 294         $line = str_replace("\t", '        ', $line);
 295
 296         parseUrl ($line);
 297
 298         $quotes = 0;
 299         $pos = 0;
 300         $j = strlen($line);
 301
 302         while ($pos < $j) {
 303             if ($line[$pos] == ' ') {
 304                 $pos++;
 305             } else if (strpos($line, '&gt;', $pos) === $pos) {
 306                 $pos += 4;
 307                 $quotes++;
 308             } else {
 309                 break;
 310             }
 311         }
 312
 313         if ($quotes % 2) {
 314             if (!isset($color[13])) {
 315                 $color[13] = '#800000';
 316             }
 317             $line = '<font color="' . $color[13] . '">' . $line . '</font>';
 318         } elseif ($quotes) {
 319             if (!isset($color[14])) {
 320                 $color[14] = '#FF0000';
 321             }
 322             $line = '<font color="' . $color[14] . '">' . $line . '</font>';
 323         }
 324
 325         $body_ary[$i] = $line;
 326     }
 327     $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
 328 }
 329
 330 /* This returns a parsed string called $body. That string can then
 331  * be displayed as the actual message in the HTML. It contains
 332  * everything needed, including HTML Tags, Attachments at the
 333  * bottom, etc.
 334  */
 335 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
 336     /* This if statement checks for the entity to show as the
 337      * primary message. To add more of them, just put them in the
 338      * order that is their priority.
 339      */
 340     global $startMessage, $username, $key, $imapServerAddress, $imapPort,
 341            $show_html_default, $sort, $has_unsafe_images, $passed_ent_id;
 342     global $languages, $squirrelmail_language;
 343
 344     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
 345         $view_unsafe_images = false;
 346     }
 347
 348     $body = '';
 349     $urlmailbox = urlencode($mailbox);
 350     $body_message = getEntity($message, $ent_num);
 351     if (($body_message->header->type0 == 'text') ||
 352         ($body_message->header->type0 == 'rfc822')) {
 353         $body = mime_fetch_body ($imap_stream, $id, $ent_num);
 354         $body = decodeBody($body, $body_message->header->encoding);
 355
 356         if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 357             function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 358             if (mb_detect_encoding($body) != 'ASCII') {
 359                 $body = $languages[$squirrelmail_language]['XTRA_CODE']('decode', $body);
 360             }
 361         }
 362         $hookResults = do_hook("message_body", $body);
 363         $body = $hookResults[1];
 364
 365         /* If there are other types that shouldn't be formatted, add
 366          * them here.
 367          */
 368
 369         if ($body_message->header->type1 == 'html') {
 370             if ($show_html_default <> 1) {
 371                 $entity_conv = array('&nbsp;' => ' ',
 372                                      '<p>'    => "\n",
 373                                      '<br>'   => "\n",
 374                                      '<P>'    => "\n",
 375                                      '<BR>'   => "\n",
 376                                      '&gt;'   => '>',
 377                                      '&lt;'   => '<');
 378                 $body = strtr($body, $entity_conv);
 379                 $body = strip_tags($body);
 380                 $body = trim($body);
 381                 translateText($body, $wrap_at,
 382                               $body_message->header->getParameter('charset'));
 383             } else {
 384                 $body = magicHTML($body, $id, $message, $mailbox);
 385             }
 386         } else {
 387             translateText($body, $wrap_at,
 388                           $body_message->header->getParameter('charset'));
 389         }
 390         $link = 'read_body.php?passed_id=' . $id . '&amp;ent_id='.$ent_num.
 391                 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
 392                 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
 393         if (isset($passed_ent_id)) {
 394             $link .= '&amp;passed_ent_id='.$passed_ent_id;
 395         }
 396         if ($view_unsafe_images) {
 397             $text = _("Hide Unsafe Images");
 398         } else {
 399             if (isset($has_unsafe_images) && $has_unsafe_images) {
 400                 $link .= '&amp;view_unsafe_images=1';
 401                 $text = _("View Unsafe Images");
 402             } else {
 403                 $text = '';
 404             }
 405         }
 406         $body .= '<center><small><a href="'.$link.'">'.$text.
 407                  '</a></small></center><br>' . "\n";
 408     }
 409     return $body;
 410 }
 411
 412
 413 function formatAttachments($message, $exclude_id, $mailbox, $id) {
 414     global $where, $what, $startMessage, $color;
 415     static $ShownHTML = 0;
 416
 417     $att_ar = $message->getAttachments($exclude_id);
 418
 419     if (!count($att_ar)) return '';
 420
 421     $attachments = '';
 422
 423     $urlMailbox = urlencode($mailbox);
 424
 425     foreach ($att_ar as $att) {
 426         $ent = $att->entity_id;
 427         $header = $att->header;
 428         $type0 = strtolower($header->type0);
 429         $type1 = strtolower($header->type1);
 430         $name = '';
 431         $links['download link']['text'] = _("download");
 432         $links['download link']['href'] = SM_PATH .
 433                 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
 434         $ImageURL = '';
 435         if ($type0 =='message' && $type1 == 'rfc822') {
 436             $default_page = SM_PATH . 'src/read_body.php';
 437             $rfc822_header = $att->rfc822_header;
 438             $filename = $rfc822_header->subject;
 439             if (trim( $filename ) == '') {
 440                 $filename = 'untitled-[' . $ent . ']' ;
 441             }
 442             $from_o = $rfc822_header->from;
 443             if (is_object($from_o)) {
 444                 $from_name = decodeHeader($from_o->getAddress(false));
 445             } else {
 446                 $from_name = _("Unknown sender");
 447             }
 448             $description = $from_name;
 449         } else {
 450             $default_page = SM_PATH . 'src/download.php';
 451             if (is_object($header->disposition)) {
 452                 $filename = $header->disposition->getProperty('filename');
 453                 if (trim($filename) == '') {
 454                     $name = decodeHeader($header->disposition->getProperty('name'));
 455                     if (trim($name) == '') {
 456                         $name = $header->getParameter('name');
 457                         if(trim($name) == '') {
 458                             if (trim( $header->id ) == '') {
 459                                 $filename = 'untitled-[' . $ent . ']' ;
 460                             } else {
 461                                 $filename = 'cid: ' . $header->id;
 462                             }
 463                         } else {
 464                             $filename = $name;
 465                         }
 466                     } else {
 467                         $filename = $name;
 468                     }
 469                 }
 470             } else {
 471                 $filename = $header->getParameter('name');
 472                 if (!trim($filename)) {
 473                     if (trim( $header->id ) == '') {
 474                         $filename = 'untitled-[' . $ent . ']' ;
 475                     } else {
 476                         $filename = 'cid: ' . $header->id;
 477                     }
 478                 }
 479             }
 480             if ($header->description) {
 481                 $description = decodeHeader($header->description);
 482             } else {
 483                 $description = '';
 484             }
 485         }
 486
 487         $display_filename = $filename;
 488         if (isset($passed_ent_id)) {
 489             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 490         } else {
 491             $passed_ent_id_link = '';
 492         }
 493         $defaultlink = $default_page . "?startMessage=$startMessage"
 494                      . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 495                      . '&amp;ent_id='.$ent.$passed_ent_id_link;
 496         if ($where && $what) {
 497            $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 498         }
 499         /* This executes the attachment hook with a specific MIME-type.
 500          * If that doesn't have results, it tries if there's a rule
 501          * for a more generic type.
 502          */
 503         $hookresults = do_hook("attachment $type0/$type1", $links,
 504                                $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 505                                $display_filename, $where, $what);
 506         if(count($hookresults[1]) <= 1) {
 507             $hookresults = do_hook("attachment $type0/*", $links,
 508                                    $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 509                                    $display_filename, $where, $what);
 510         }
 511
 512         $links = $hookresults[1];
 513         $defaultlink = $hookresults[6];
 514
 515         $attachments .= '<TR><TD>' .
 516                         '<A HREF="'.$defaultlink.'">'.decodeHeader($display_filename).'</A>&nbsp;</TD>' .
 517                         '<TD><SMALL><b>' . show_readable_size($header->size) .
 518                         '</b>&nbsp;&nbsp;</small></TD>' .
 519                         "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
 520                         '<TD><SMALL>';
 521         $attachments .= '<b>' . $description . '</b>';
 522         $attachments .= '</SMALL></TD><TD><SMALL>&nbsp;';
 523
 524         $skipspaces = 1;
 525         foreach ($links as $val) {
 526             if ($skipspaces) {
 527                 $skipspaces = 0;
 528             } else {
 529                 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
 530             }
 531             $attachments .= '<a href="' . $val['href'] . '">' .  $val['text'] . '</a>';
 532         }
 533         unset($links);
 534         $attachments .= "</TD></TR>\n";
 535     }
 536     $attachmentadd = do_hook_function('attachments_bottom',$attachments);
 537     if ($attachmentadd != '')
 538         $attachments = $attachmentadd;
 539     return $attachments;
 540 }
 541
 542 function sqimap_base64_decode(&$string) {
 543     $string = str_replace("\r\n", "\n", $string);
 544     $string = base64_decode($string);
 545 }
 546
 547 /* This function decodes the body depending on the encoding type. */
 548 function decodeBody($body, $encoding) {
 549     global $show_html_default;
 550
 551     $body = str_replace("\r\n", "\n", $body);
 552     $encoding = strtolower($encoding);
 553
 554     $encoding_handler = do_hook_function('decode_body', $encoding);
 555
 556
 557     // plugins get first shot at decoding the body
 558     //
 559     if (!empty($encoding_handler) && function_exists($encoding_handler)) {
 560         $body = $encoding_handler('decode', $body);
 561
 562     } else if ($encoding == 'quoted-printable' ||
 563         $encoding == 'quoted_printable') {
 564         $body = quoted_printable_decode($body);
 565
 566         while (ereg("=\n", $body)) {
 567             $body = ereg_replace ("=\n", '', $body);
 568         }
 569
 570     } else if ($encoding == 'base64') {
 571         $body = base64_decode($body);
 572     }
 573
 574     // All other encodings are returned raw.
 575     return $body;
 576 }
 577
 578 /*
 579  * This functions decode strings that is encoded according to
 580  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 581  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 582  */
 583 function decodeHeader ($string, $utfencode=true,$htmlsave=true) {
 584     global $languages, $squirrelmail_language;
 585     if (is_array($string)) {
 586         $string = implode("\n", $string);
 587     }
 588
 589     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 590         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 591         $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
 592         // Do we need to return at this point?
 593         // return $string;
 594     }
 595     $i = 0;
 596     $iLastMatch = -2;
 597     $encoded = true;
 598
 599     $aString = explode(' ',$string);
 600     $ret = '';
 601     foreach ($aString as $chunk) {
 602         if ($encoded && $chunk === '') {
 603             continue;
 604         } elseif ($chunk === '') {
 605             $ret .= ' ';
 606             continue;
 607         }
 608         $encoded = false;
 609         /* if encoded words are not separated by a linear-space-white we still catch them */
 610         $j = $i-1;
 611 //      if ($chunk{0} === '=') { /* performance, saves an unnessecarry preg call */
 612         while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 613             /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
 614             if ($iLastMatch !== $j) {
 615                 if ($htmlsave) {
 616                     $ret .= '&nbsp;';
 617                 } else {
 618                     $ret .= ' ';
 619                 }
 620             }
 621             $iLastMatch = $i;
 622             $j = $i;
 623             $ret .= $res[1];
 624             $encoding = ucfirst($res[3]);
 625             switch ($encoding)
 626             {
 627             case 'B':
 628                 $replace = base64_decode($res[4]);
 629                 $ret .= charset_decode($res[2],$replace);
 630                 break;
 631             case 'Q':
 632                 $replace = str_replace('_', ' ', $res[4]);
 633                 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 634                                     $replace);
 635                 /* Only encode into entities by default. Some places
 636                  * don't need the encoding, like the compose form.
 637                  */
 638                 if ($utfencode) {
 639                     $replace = charset_decode($res[2], $replace);
 640                 } else {
 641                     if ($htmlsave) {
 642                         $replace = htmlspecialchars($replace);
 643                     }
 644                 }
 645                 $ret .= $replace;
 646                 break;
 647             default:
 648                 break;
 649             }
 650             $chunk = $res[5];
 651             $encoded = true;
 652         }
 653 //      }
 654         if (!$encoded) {
 655             if ($htmlsave) {
 656                 $ret .= '&nbsp;';
 657             } else {
 658                 $ret .= ' ';
 659             }
 660         }
 661
 662         if (!$encoded && $htmlsave) {
 663             $ret .= htmlspecialchars($chunk);
 664         } else {
 665             $ret .= $chunk;
 666         }
 667         ++$i;
 668     }
 669     /* remove the first added space */
 670     if ($ret) {
 671         if ($htmlsave) {
 672             $ret = substr($ret,6);
 673         } else {
 674             $ret = substr($ret,1);
 675         }
 676     }
 677
 678     return $ret;
 679 }
 680
 681 /*
 682  * Encode a string according to RFC 1522 for use in headers if it
 683  * contains 8-bit characters or anything that looks like it should
 684  * be encoded.
 685  */
 686 function encodeHeader ($string) {
 687     global $default_charset, $languages, $squirrelmail_language;
 688
 689     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 690         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 691         return  $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
 692     }
 693     if (strtolower($default_charset) == 'iso-8859-1') {
 694         $string = str_replace("\240",' ',$string);
 695     }
 696
 697     // Encode only if the string contains 8-bit characters or =?
 698     $j = strlen($string);
 699     $max_l = 75 - strlen($default_charset) - 7;
 700     $aRet = array();
 701     $ret = '';
 702     $iEncStart = $enc_init = false;
 703     $cur_l = $iOffset = 0;
 704     for($i = 0; $i < $j; ++$i) {
 705         switch($string{$i})
 706         {
 707         case '=':
 708         case '<':
 709         case '>':
 710         case ',':
 711         case '?':
 712         case '_':
 713             if ($iEncStart === false) {
 714                 $iEncStart = $i;
 715             }
 716             $cur_l+=3;
 717             if ($cur_l > ($max_l-2)) {
 718                 /* if there is an stringpart that doesn't need encoding, add it */
 719                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 720                 $aRet[] = "=?$default_charset?Q?$ret?=";
 721                 $iOffset = $i;
 722                 $cur_l = 0;
 723                 $ret = '';
 724                 $iEncStart = false;
 725             } else {
 726                 $ret .= sprintf("=%02X",ord($string{$i}));
 727             }
 728             break;
 729         case '(':
 730         case ')':
 731             if ($iEncStart !== false) {
 732                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 733                 $aRet[] = "=?$default_charset?Q?$ret?=";
 734                 $iOffset = $i;
 735                 $cur_l = 0;
 736                 $ret = '';
 737                 $iEncStart = false;
 738             }
 739             break;
 740         case ' ':
 741             if ($iEncStart !== false) {
 742                 $cur_l++;
 743                 if ($cur_l > $max_l) {
 744                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 745                     $aRet[] = "=?$default_charset?Q?$ret?=";
 746                     $iOffset = $i;
 747                     $cur_l = 0;
 748                     $ret = '';
 749                     $iEncStart = false;
 750                 } else {
 751                     $ret .= '_';
 752                 }
 753             }
 754             break;
 755         default:
 756             $k = ord($string{$i});
 757             if ($k > 126) {
 758                 if ($iEncStart === false) {
 759                     // do not start encoding in the middle of a string, also take the rest of the word.
 760                     $sLeadString = substr($string,0,$i);
 761                     $aLeadString = explode(' ',$sLeadString);
 762                     $sToBeEncoded = array_pop($aLeadString);
 763                     $iEncStart = $i - strlen($sToBeEncoded);
 764                     $ret .= $sToBeEncoded;
 765                     $cur_l += strlen($sToBeEncoded);
 766                 }
 767                 $cur_l += 3;
 768                 /* first we add the encoded string that reached it's max size */
 769                 if ($cur_l > ($max_l-2)) {
 770                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 771                     $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
 772                     $cur_l = 3;
 773                     $ret = '';
 774                     $iOffset = $i;
 775                     $iEncStart = $i;
 776                 }
 777                 $enc_init = true;
 778                 $ret .= sprintf("=%02X", $k);
 779             } else {
 780                 if ($iEncStart !== false) {
 781                     $cur_l++;
 782                     if ($cur_l > $max_l) {
 783                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 784                         $aRet[] = "=?$default_charset?Q?$ret?=";
 785                         $iEncStart = false;
 786                         $iOffset = $i;
 787                         $cur_l = 0;
 788                         $ret = '';
 789                     } else {
 790                         $ret .= $string{$i};
 791                     }
 792                 }
 793             }
 794             break;
 795         }
 796     }
 797
 798     if ($enc_init) {
 799         if ($iEncStart !== false) {
 800             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 801             $aRet[] = "=?$default_charset?Q?$ret?=";
 802         } else {
 803             $aRet[] = substr($string,$iOffset);
 804         }
 805         $string = implode('',$aRet);
 806     }
 807     return $string;
 808 }
 809
 810 /* This function trys to locate the entity_id of a specific mime element */
 811 function find_ent_id($id, $message) {
 812     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
 813         if ($message->entities[$i]->header->type0 == 'multipart')  {
 814             $ret = find_ent_id($id, $message->entities[$i]);
 815         } else {
 816             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
 817 //                if (sq_check_save_extension($message->entities[$i])) {
 818                     return $message->entities[$i]->entity_id;
 819 //                }
 820             }
 821         }
 822     }
 823     return $ret;
 824 }
 825
 826 function sq_check_save_extension($message) {
 827     $filename = $message->getFilename();
 828     $ext = substr($filename, strrpos($filename,'.')+1);
 829     $save_extensions = array('jpg','jpeg','gif','png','bmp');
 830     return in_array($ext, $save_extensions);
 831 }
 832
 833
 834 /**
 835  ** HTMLFILTER ROUTINES
 836  */
 837
 838 /**
 839  * This function is more or less a wrapper around stripslashes. Apparently
 840  * Explorer is stupid enough to just remove the backslashes and then
 841  * execute the content of the attribute as if nothing happened.
 842  * Who does that?
 843  *
 844  * @param  attvalue   The value of the attribute
 845  * @return attvalue   The value of the attribute stripslashed.
 846  */
 847 function sq_unbackslash($attvalue){
 848     /**
 849      * Remove any backslashes. See if there are any first.
 850      */
 851     if (strstr($attvalue, '\\') !== false){
 852         $attvalue = stripslashes($attvalue);
 853     }
 854     return $attvalue;
 855 }
 856
 857 /**
 858  * Kill any tabs, newlines, or carriage returns. Our friends the
 859  * makers of the browser with 95% market value decided that it'd
 860  * be funny to make "java[tab]script" be just as good as "javascript".
 861  *
 862  * @param  attvalue  The attribute value before extraneous spaces removed.
 863  * @return attvalue  The attribute value after extraneous spaces removed.
 864  */
 865 function sq_unspace($attvalue){
 866     if (strcspn($attvalue, "\t\r\n") != strlen($attvalue)){
 867         $attvalue = str_replace(Array("\t", "\r", "\n"), Array('', '', ''),
 868                                 $attvalue);
 869     }
 870     return $attvalue;
 871 }
 872
 873 /**
 874  * This function returns the final tag out of the tag name, an array
 875  * of attributes, and the type of the tag. This function is called by
 876  * sq_sanitize internally.
 877  *
 878  * @param  $tagname  the name of the tag.
 879  * @param  $attary   the array of attributes and their values
 880  * @param  $tagtype  The type of the tag (see in comments).
 881  * @return           a string with the final tag representation.
 882  */
 883 function sq_tagprint($tagname, $attary, $tagtype){
 884     $me = 'sq_tagprint';
 885
 886     if ($tagtype == 2){
 887         $fulltag = '</' . $tagname . '>';
 888     } else {
 889         $fulltag = '<' . $tagname;
 890         if (is_array($attary) && sizeof($attary)){
 891             $atts = Array();
 892             while (list($attname, $attvalue) = each($attary)){
 893                 array_push($atts, "$attname=$attvalue");
 894             }
 895             $fulltag .= ' ' . join(" ", $atts);
 896         }
 897         if ($tagtype == 3){
 898             $fulltag .= ' /';
 899         }
 900         $fulltag .= '>';
 901     }
 902     return $fulltag;
 903 }
 904
 905 /**
 906  * A small helper function to use with array_walk. Modifies a by-ref
 907  * value and makes it lowercase.
 908  *
 909  * @param  $val a value passed by-ref.
 910  * @return      void since it modifies a by-ref value.
 911  */
 912 function sq_casenormalize(&$val){
 913     $val = strtolower($val);
 914 }
 915
 916 /**
 917  * This function skips any whitespace from the current position within
 918  * a string and to the next non-whitespace value.
 919  *
 920  * @param  $body   the string
 921  * @param  $offset the offset within the string where we should start
 922  *                 looking for the next non-whitespace character.
 923  * @return         the location within the $body where the next
 924  *                 non-whitespace char is located.
 925  */
 926 function sq_skipspace($body, $offset){
 927     $me = 'sq_skipspace';
 928     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
 929     if (sizeof($matches{1})){
 930         $count = strlen($matches{1});
 931         $offset += $count;
 932     }
 933     return $offset;
 934 }
 935
 936 /**
 937  * This function looks for the next character within a string.  It's
 938  * really just a glorified "strpos", except it catches if failures
 939  * nicely.
 940  *
 941  * @param  $body   The string to look for needle in.
 942  * @param  $offset Start looking from this position.
 943  * @param  $needle The character/string to look for.
 944  * @return         location of the next occurance of the needle, or
 945  *                 strlen($body) if needle wasn't found.
 946  */
 947 function sq_findnxstr($body, $offset, $needle){
 948     $me  = 'sq_findnxstr';
 949     $pos = strpos($body, $needle, $offset);
 950     if ($pos === FALSE){
 951         $pos = strlen($body);
 952     }
 953     return $pos;
 954 }
 955
 956 /**
 957  * This function takes a PCRE-style regexp and tries to match it
 958  * within the string.
 959  *
 960  * @param  $body   The string to look for needle in.
 961  * @param  $offset Start looking from here.
 962  * @param  $reg    A PCRE-style regex to match.
 963  * @return         Returns a false if no matches found, or an array
 964  *                 with the following members:
 965  *                 - integer with the location of the match within $body
 966  *                 - string with whatever content between offset and the match
 967  *                 - string with whatever it is we matched
 968  */
 969 function sq_findnxreg($body, $offset, $reg){
 970     $me = 'sq_findnxreg';
 971     $matches = Array();
 972     $retarr = Array();
 973     preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
 974     if (!isset($matches{0}) || !$matches{0}){
 975         $retarr = false;
 976     } else {
 977         $retarr{0} = $offset + strlen($matches{1});
 978         $retarr{1} = $matches{1};
 979         $retarr{2} = $matches{2};
 980     }
 981     return $retarr;
 982 }
 983
 984 /**
 985  * This function looks for the next tag.
 986  *
 987  * @param  $body   String where to look for the next tag.
 988  * @param  $offset Start looking from here.
 989  * @return         false if no more tags exist in the body, or
 990  *                 an array with the following members:
 991  *                 - string with the name of the tag
 992  *                 - array with attributes and their values
 993  *                 - integer with tag type (1, 2, or 3)
 994  *                 - integer where the tag starts (starting "<")
 995  *                 - integer where the tag ends (ending ">")
 996  *                 first three members will be false, if the tag is invalid.
 997  */
 998 function sq_getnxtag($body, $offset){
 999     $me = 'sq_getnxtag';
1000     if ($offset > strlen($body)){
1001         return false;
1002     }
1003     $lt = sq_findnxstr($body, $offset, "<");
1004     if ($lt == strlen($body)){
1005         return false;
1006     }
1007     /**
1008      * We are here:
1009      * blah blah <tag attribute="value">
1010      * \---------^
1011      */
1012     $pos = sq_skipspace($body, $lt+1);
1013     if ($pos >= strlen($body)){
1014         return Array(false, false, false, $lt, strlen($body));
1015     }
1016     /**
1017      * There are 3 kinds of tags:
1018      * 1. Opening tag, e.g.:
1019      *    <a href="blah">
1020      * 2. Closing tag, e.g.:
1021      *    </a>
1022      * 3. XHTML-style content-less tag, e.g.:
1023      *    <img src="blah"/>
1024      */
1025     $tagtype = false;
1026     switch (substr($body, $pos, 1)){
1027         case '/':
1028             $tagtype = 2;
1029             $pos++;
1030             break;
1031         case '!':
1032             /**
1033              * A comment or an SGML declaration.
1034              */
1035             if (substr($body, $pos+1, 2) == "--"){
1036                 $gt = strpos($body, "-->", $pos);
1037                 if ($gt === false){
1038                     $gt = strlen($body);
1039                 } else {
1040                     $gt += 2;
1041                 }
1042                 return Array(false, false, false, $lt, $gt);
1043             } else {
1044                 $gt = sq_findnxstr($body, $pos, ">");
1045                 return Array(false, false, false, $lt, $gt);
1046             }
1047             break;
1048         default:
1049             /**
1050              * Assume tagtype 1 for now. If it's type 3, we'll switch values
1051              * later.
1052              */
1053             $tagtype = 1;
1054             break;
1055     }
1056
1057     $tag_start = $pos;
1058     $tagname = '';
1059     /**
1060      * Look for next [\W-_], which will indicate the end of the tag name.
1061      */
1062     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1063     if ($regary == false){
1064         return Array(false, false, false, $lt, strlen($body));
1065     }
1066     list($pos, $tagname, $match) = $regary;
1067     $tagname = strtolower($tagname);
1068
1069     /**
1070      * $match can be either of these:
1071      * '>'  indicating the end of the tag entirely.
1072      * '\s' indicating the end of the tag name.
1073      * '/'  indicating that this is type-3 xhtml tag.
1074      *
1075      * Whatever else we find there indicates an invalid tag.
1076      */
1077     switch ($match){
1078         case '/':
1079             /**
1080              * This is an xhtml-style tag with a closing / at the
1081              * end, like so: <img src="blah"/>. Check if it's followed
1082              * by the closing bracket. If not, then this tag is invalid
1083              */
1084             if (substr($body, $pos, 2) == "/>"){
1085                 $pos++;
1086                 $tagtype = 3;
1087             } else {
1088                 $gt = sq_findnxstr($body, $pos, ">");
1089                 $retary = Array(false, false, false, $lt, $gt);
1090                 return $retary;
1091             }
1092         case '>':
1093             return Array($tagname, false, $tagtype, $lt, $pos);
1094             break;
1095         default:
1096             /**
1097              * Check if it's whitespace
1098              */
1099             if (!preg_match('/\s/', $match)){
1100                 /**
1101                  * This is an invalid tag! Look for the next closing ">".
1102                  */
1103                 $gt = sq_findnxstr($body, $lt, ">");
1104                 return Array(false, false, false, $lt, $gt);
1105             }
1106             break;
1107     }
1108
1109     /**
1110      * At this point we're here:
1111      * <tagname  attribute='blah'>
1112      * \-------^
1113      *
1114      * At this point we loop in order to find all attributes.
1115      */
1116     $attname = '';
1117     $atttype = false;
1118     $attary = Array();
1119
1120     while ($pos <= strlen($body)){
1121         $pos = sq_skipspace($body, $pos);
1122         if ($pos == strlen($body)){
1123             /**
1124              * Non-closed tag.
1125              */
1126             return Array(false, false, false, $lt, $pos);
1127         }
1128         /**
1129          * See if we arrived at a ">" or "/>", which means that we reached
1130          * the end of the tag.
1131          */
1132         $matches = Array();
1133         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1134             /**
1135              * Yep. So we did.
1136              */
1137             $pos += strlen($matches{1});
1138             if ($matches{2} == "/>"){
1139                 $tagtype = 3;
1140                 $pos++;
1141             }
1142             return Array($tagname, $attary, $tagtype, $lt, $pos);
1143         }
1144
1145         /**
1146          * There are several types of attributes, with optional
1147          * [:space:] between members.
1148          * Type 1:
1149          *   attrname[:space:]=[:space:]'CDATA'
1150          * Type 2:
1151          *   attrname[:space:]=[:space:]"CDATA"
1152          * Type 3:
1153          *   attr[:space:]=[:space:]CDATA
1154          * Type 4:
1155          *   attrname
1156          *
1157          * We leave types 1 and 2 the same, type 3 we check for
1158          * '"' and convert to "&quot" if needed, then wrap in
1159          * double quotes. Type 4 we convert into:
1160          * attrname="yes".
1161          */
1162         $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1163         if ($regary == false){
1164             /**
1165              * Looks like body ended before the end of tag.
1166              */
1167             return Array(false, false, false, $lt, strlen($body));
1168         }
1169         list($pos, $attname, $match) = $regary;
1170         $attname = strtolower($attname);
1171         /**
1172          * We arrived at the end of attribute name. Several things possible
1173          * here:
1174          * '>'  means the end of the tag and this is attribute type 4
1175          * '/'  if followed by '>' means the same thing as above
1176          * '\s' means a lot of things -- look what it's followed by.
1177          *      anything else means the attribute is invalid.
1178          */
1179         switch($match){
1180             case '/':
1181                 /**
1182                  * This is an xhtml-style tag with a closing / at the
1183                  * end, like so: <img src="blah"/>. Check if it's followed
1184                  * by the closing bracket. If not, then this tag is invalid
1185                  */
1186                 if (substr($body, $pos, 2) == "/>"){
1187                     $pos++;
1188                     $tagtype = 3;
1189                 } else {
1190                     $gt = sq_findnxstr($body, $pos, ">");
1191                     $retary = Array(false, false, false, $lt, $gt);
1192                     return $retary;
1193                 }
1194             case '>':
1195                 $attary{$attname} = '"yes"';
1196                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1197                 break;
1198             default:
1199                 /**
1200                  * Skip whitespace and see what we arrive at.
1201                  */
1202                 $pos = sq_skipspace($body, $pos);
1203                 $char = substr($body, $pos, 1);
1204                 /**
1205                  * Two things are valid here:
1206                  * '=' means this is attribute type 1 2 or 3.
1207                  * \w means this was attribute type 4.
1208                  * anything else we ignore and re-loop. End of tag and
1209                  * invalid stuff will be caught by our checks at the beginning
1210                  * of the loop.
1211                  */
1212                 if ($char == "="){
1213                     $pos++;
1214                     $pos = sq_skipspace($body, $pos);
1215                     /**
1216                      * Here are 3 possibilities:
1217                      * "'"  attribute type 1
1218                      * '"'  attribute type 2
1219                      * everything else is the content of tag type 3
1220                      */
1221                     $quot = substr($body, $pos, 1);
1222                     if ($quot == "'"){
1223                         $regary = sq_findnxreg($body, $pos+1, "\'");
1224                         if ($regary == false){
1225                             return Array(false, false, false, $lt, strlen($body));
1226                         }
1227                         list($pos, $attval, $match) = $regary;
1228                         $pos++;
1229                         $attary{$attname} = "'" . $attval . "'";
1230                     } else if ($quot == '"'){
1231                         $regary = sq_findnxreg($body, $pos+1, '\"');
1232                         if ($regary == false){
1233                             return Array(false, false, false, $lt, strlen($body));
1234                         }
1235                         list($pos, $attval, $match) = $regary;
1236                         $pos++;
1237                         $attary{$attname} = '"' . $attval . '"';
1238                     } else {
1239                         /**
1240                          * These are hateful. Look for \s, or >.
1241                          */
1242                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1243                         if ($regary == false){
1244                             return Array(false, false, false, $lt, strlen($body));
1245                         }
1246                         list($pos, $attval, $match) = $regary;
1247                         /**
1248                          * If it's ">" it will be caught at the top.
1249                          */
1250                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1251                         $attary{$attname} = '"' . $attval . '"';
1252                     }
1253                 } else if (preg_match("|[\w/>]|", $char)) {
1254                     /**
1255                      * That was attribute type 4.
1256                      */
1257                     $attary{$attname} = '"yes"';
1258                 } else {
1259                     /**
1260                      * An illegal character. Find next '>' and return.
1261                      */
1262                     $gt = sq_findnxstr($body, $pos, ">");
1263                     return Array(false, false, false, $lt, $gt);
1264                 }
1265                 break;
1266         }
1267     }
1268     /**
1269      * The fact that we got here indicates that the tag end was never
1270      * found. Return invalid tag indication so it gets stripped.
1271      */
1272     return Array(false, false, false, $lt, strlen($body));
1273 }
1274
1275 /**
1276  * This function checks attribute values for entity-encoded values
1277  * and returns them translated into 8-bit strings so we can run
1278  * checks on them.
1279  *
1280  * @param  $attvalue A string to run entity check against.
1281  * @return           Translated value.
1282  */
1283 function sq_deent($attvalue){
1284     $me = 'sq_deent';
1285     /**
1286      * See if we have to run the checks first. All entities must start
1287      * with "&".
1288      */
1289     if (strpos($attvalue, "&") === false){
1290         return $attvalue;
1291     }
1292     /**
1293      * Check named entities first.
1294      */
1295     $trans = get_html_translation_table(HTML_ENTITIES);
1296     /**
1297      * Leave &quot; in, as it can mess us up.
1298      */
1299     $trans = array_flip($trans);
1300     unset($trans{"&quot;"});
1301     while (list($ent, $val) = each($trans)){
1302         $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1303     }
1304     /**
1305      * Now translate numbered entities from 1 to 255 if needed.
1306      */
1307     if (strpos($attvalue, "#") !== false){
1308         $omit = Array(34, 39);
1309         for ($asc=1; $asc<256; $asc++){
1310             if (!in_array($asc, $omit)){
1311                 $chr = chr($asc);
1312                 $attvalue = preg_replace("/\&#0*$asc;*(\D)/si", "$chr\\1",
1313                                          $attvalue);
1314                 $attvalue = preg_replace("/\&#x0*".dechex($asc).";*(\W)/si",
1315                                          "$chr\\1", $attvalue);
1316             }
1317         }
1318     }
1319     return $attvalue;
1320 }
1321
1322 /**
1323  * This function runs various checks against the attributes.
1324  *
1325  * @param  $tagname         String with the name of the tag.
1326  * @param  $attary          Array with all tag attributes.
1327  * @param  $rm_attnames     See description for sq_sanitize
1328  * @param  $bad_attvals     See description for sq_sanitize
1329  * @param  $add_attr_to_tag See description for sq_sanitize
1330  * @param  $message         message object
1331  * @param  $id              message id
1332  * @return                  Array with modified attributes.
1333  */
1334 function sq_fixatts($tagname,
1335                     $attary,
1336                     $rm_attnames,
1337                     $bad_attvals,
1338                     $add_attr_to_tag,
1339                     $message,
1340                     $id,
1341                     $mailbox
1342                     ){
1343     $me = 'sq_fixatts';
1344     while (list($attname, $attvalue) = each($attary)){
1345         /**
1346          * See if this attribute should be removed.
1347          */
1348         foreach ($rm_attnames as $matchtag=>$matchattrs){
1349             if (preg_match($matchtag, $tagname)){
1350                 foreach ($matchattrs as $matchattr){
1351                     if (preg_match($matchattr, $attname)){
1352                         unset($attary{$attname});
1353                         continue;
1354                     }
1355                 }
1356             }
1357         }
1358         /**
1359          * Remove any backslashes, entities, and extraneous whitespace.
1360          */
1361         $attvalue = sq_unbackslash($attvalue);
1362         $attvalue = sq_deent($attvalue);
1363         $attvalue = sq_unspace($attvalue);
1364
1365         /**
1366          * Now let's run checks on the attvalues.
1367          * I don't expect anyone to comprehend this. If you do,
1368          * get in touch with me so I can drive to where you live and
1369          * shake your hand personally. :)
1370          */
1371         foreach ($bad_attvals as $matchtag=>$matchattrs){
1372             if (preg_match($matchtag, $tagname)){
1373                 foreach ($matchattrs as $matchattr=>$valary){
1374                     if (preg_match($matchattr, $attname)){
1375                         /**
1376                          * There are two arrays in valary.
1377                          * First is matches.
1378                          * Second one is replacements
1379                          */
1380                         list($valmatch, $valrepl) = $valary;
1381                         $newvalue =
1382                             preg_replace($valmatch, $valrepl, $attvalue);
1383                         if ($newvalue != $attvalue){
1384                             $attary{$attname} = $newvalue;
1385                         }
1386                     }
1387                 }
1388             }
1389         }
1390         /**
1391          * Turn cid: urls into http-friendly ones.
1392          */
1393         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1394             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1395         }
1396     }
1397     /**
1398      * See if we need to append any attributes to this tag.
1399      */
1400     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1401         if (preg_match($matchtag, $tagname)){
1402             $attary = array_merge($attary, $addattary);
1403         }
1404     }
1405     return $attary;
1406 }
1407
1408 /**
1409  * This function edits the style definition to make them friendly and
1410  * usable in squirrelmail.
1411  *
1412  * @param  $message  the message object
1413  * @param  $id       the message id
1414  * @param  $content  a string with whatever is between <style> and </style>
1415  * @return           a string with edited content.
1416  */
1417 function sq_fixstyle($body, $pos, $message, $id){
1418     global $view_unsafe_images;
1419     $me = 'sq_fixstyle';
1420     $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1421     if ($ret == FALSE){
1422         return array(FALSE, strlen($body));
1423     }
1424     $newpos = $ret[0] + strlen($ret[2]);
1425     $content = $ret[1];
1426     /**
1427      * First look for general BODY style declaration, which would be
1428      * like so:
1429      * body {background: blah-blah}
1430      * and change it to .bodyclass so we can just assign it to a <div>
1431      */
1432     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1433     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1434     /**
1435      * Fix url('blah') declarations.
1436      */
1437     $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1438                             "url(\\1$secremoveimg\\2)", $content);
1439     /**
1440      * Fix url('https*://.*) declarations but only if $view_unsafe_images
1441      * is false.
1442      */
1443     if (!$view_unsafe_images){
1444         $content = preg_replace("|url\s*\(\s*([\'\"])\s*https*:.*?([\'\"])\s*\)|si",
1445                                 "url(\\1$secremoveimg\\2)", $content);
1446     }
1447
1448     /**
1449      * Fix urls that refer to cid:
1450      */
1451     while (preg_match("|url\s*\(\s*([\'\"]\s*cid:.*?[\'\"])\s*\)|si",
1452                       $content, $matches)){
1453         $cidurl = $matches{1};
1454         $httpurl = sq_cid2http($message, $id, $cidurl);
1455         $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1456                                 "url($httpurl)", $content);
1457     }
1458
1459     /**
1460      * Fix stupid css declarations which lead to vulnerabilities
1461      * in IE.
1462      */
1463     $match   = Array('/expression/i',
1464                      '/behaviou*r/i',
1465                      '/binding/i',
1466                      '/include-source/i');
1467     $replace = Array('idiocy', 'idiocy', 'idiocy', 'idiocy');
1468     $content = preg_replace($match, $replace, $content);
1469     return array($content, $newpos);
1470 }
1471
1472 /**
1473  * This function converts cid: url's into the ones that can be viewed in
1474  * the browser.
1475  *
1476  * @param  $message  the message object
1477  * @param  $id       the message id
1478  * @param  $cidurl   the cid: url.
1479  * @return           a string with a http-friendly url
1480  */
1481 function sq_cid2http($message, $id, $cidurl, $mailbox){
1482     /**
1483      * Get rid of quotes.
1484      */
1485     $quotchar = substr($cidurl, 0, 1);
1486     if ($quotchar == '"' || $quotchar == "'"){
1487         $cidurl = str_replace($quotchar, "", $cidurl);
1488     } else {
1489         $quotchar = '';
1490     }
1491     $cidurl = substr(trim($cidurl), 4);
1492     $linkurl = find_ent_id($cidurl, $message);
1493     /* in case of non-save cid links $httpurl should be replaced by a sort of
1494        unsave link image */
1495     $httpurl = '';
1496     if ($linkurl) {
1497         $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1498                    "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1499                    '&amp;ent_id=' . $linkurl . $quotchar;
1500     }
1501     return $httpurl;
1502 }
1503
1504 /**
1505  * This function changes the <body> tag into a <div> tag since we
1506  * can't really have a body-within-body.
1507  *
1508  * @param  $attary   an array of attributes and values of <body>
1509  * @param  $mailbox  mailbox we're currently reading (for cid2http)
1510  * @param  $message  current message (for cid2http)
1511  * @param  $id       current message id (for cid2http)
1512  * @return           a modified array of attributes to be set for <div>
1513  */
1514 function sq_body2div($attary, $mailbox, $message, $id){
1515     $me = 'sq_body2div';
1516     $divattary = Array('class' => "'bodyclass'");
1517     $bgcolor = '#ffffff';
1518     $text = '#000000';
1519     $styledef = '';
1520     if (is_array($attary) && sizeof($attary) > 0){
1521         foreach ($attary as $attname=>$attvalue){
1522             $quotchar = substr($attvalue, 0, 1);
1523             $attvalue = str_replace($quotchar, "", $attvalue);
1524             switch ($attname){
1525                 case 'background':
1526                     $attvalue = sq_cid2http($message, $id,
1527                                             $attvalue, $mailbox);
1528                     $styledef .= "background-image: url('$attvalue'); ";
1529                     break;
1530                 case 'bgcolor':
1531                     $styledef .= "background-color: $attvalue; ";
1532                     break;
1533                 case 'text':
1534                     $styledef .= "color: $attvalue; ";
1535                     break;
1536             }
1537         }
1538         if (strlen($styledef) > 0){
1539             $divattary{"style"} = "\"$styledef\"";
1540         }
1541     }
1542     return $divattary;
1543 }
1544
1545 /**
1546  * This is the main function and the one you should actually be calling.
1547  * There are several variables you should be aware of an which need
1548  * special description.
1549  *
1550  * Since the description is quite lengthy, see it here:
1551  * http://www.mricon.com/html/phpfilter.html
1552  *
1553  * @param $body                 the string with HTML you wish to filter
1554  * @param $tag_list             see description above
1555  * @param $rm_tags_with_content see description above
1556  * @param $self_closing_tags    see description above
1557  * @param $force_tag_closing    see description above
1558  * @param $rm_attnames          see description above
1559  * @param $bad_attvals          see description above
1560  * @param $add_attr_to_tag      see description above
1561  * @param $message              message object
1562  * @param $id                   message id
1563  * @return                      sanitized html safe to show on your pages.
1564  */
1565 function sq_sanitize($body,
1566                      $tag_list,
1567                      $rm_tags_with_content,
1568                      $self_closing_tags,
1569                      $force_tag_closing,
1570                      $rm_attnames,
1571                      $bad_attvals,
1572                      $add_attr_to_tag,
1573                      $message,
1574                      $id,
1575                      $mailbox
1576                      ){
1577     $me = 'sq_sanitize';
1578     $rm_tags = array_shift($tag_list);
1579     /**
1580      * Normalize rm_tags and rm_tags_with_content.
1581      */
1582     @array_walk($tag_list, 'sq_casenormalize');
1583     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1584     @array_walk($self_closing_tags, 'sq_casenormalize');
1585     /**
1586      * See if tag_list is of tags to remove or tags to allow.
1587      * false  means remove these tags
1588      * true   means allow these tags
1589      */
1590     $curpos = 0;
1591     $open_tags = Array();
1592     $trusted = "\n<!-- begin sanitized html -->\n";
1593     $skip_content = false;
1594     /**
1595      * Take care of netscape's stupid javascript entities like
1596      * &{alert('boo')};
1597      */
1598     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1599
1600     while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1601         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1602         $free_content = substr($body, $curpos, $lt-$curpos);
1603         /**
1604          * Take care of <style>
1605          */
1606         if ($tagname == "style" && $tagtype == 1){
1607             list($free_content, $curpos) =
1608                 sq_fixstyle($body, $gt+1, $message, $id);
1609             if ($free_content != FALSE){
1610                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1611                 $trusted .= $free_content;
1612                 $trusted .= sq_tagprint($tagname, false, 2);
1613             }
1614             continue;
1615         }
1616         if ($skip_content == false){
1617             $trusted .= $free_content;
1618         }
1619         if ($tagname != FALSE){
1620             if ($tagtype == 2){
1621                 if ($skip_content == $tagname){
1622                     /**
1623                      * Got to the end of tag we needed to remove.
1624                      */
1625                     $tagname = false;
1626                     $skip_content = false;
1627                 } else {
1628                     if ($skip_content == false){
1629                         if ($tagname == "body"){
1630                             $tagname = "div";
1631                         }
1632                         if (isset($open_tags{$tagname}) &&
1633                             $open_tags{$tagname} > 0){
1634                             $open_tags{$tagname}--;
1635                         } else {
1636                             $tagname = false;
1637                         }
1638                     }
1639                 }
1640             } else {
1641                 /**
1642                  * $rm_tags_with_content
1643                  */
1644                 if ($skip_content == false){
1645                     /**
1646                      * See if this is a self-closing type and change
1647                      * tagtype appropriately.
1648                      */
1649                     if ($tagtype == 1
1650                         && in_array($tagname, $self_closing_tags)){
1651                         $tagtype = 3;
1652                     }
1653                     /**
1654                      * See if we should skip this tag and any content
1655                      * inside it.
1656                      */
1657                     if ($tagtype == 1 &&
1658                         in_array($tagname, $rm_tags_with_content)){
1659                         $skip_content = $tagname;
1660                     } else {
1661                         if (($rm_tags == false
1662                              && in_array($tagname, $tag_list)) ||
1663                             ($rm_tags == true &&
1664                              !in_array($tagname, $tag_list))){
1665                             $tagname = false;
1666                         } else {
1667                             /**
1668                              * Convert body into div.
1669                              */
1670                             if ($tagname == "body"){
1671                                 $tagname = "div";
1672                                 $attary = sq_body2div($attary, $mailbox,
1673                                                       $message, $id);
1674                             }
1675                             if ($tagtype == 1){
1676                                 if (isset($open_tags{$tagname})){
1677                                     $open_tags{$tagname}++;
1678                                 } else {
1679                                     $open_tags{$tagname}=1;
1680                                 }
1681                             }
1682                             /**
1683                              * This is where we run other checks.
1684                              */
1685                             if (is_array($attary) && sizeof($attary) > 0){
1686                                 $attary = sq_fixatts($tagname,
1687                                                      $attary,
1688                                                      $rm_attnames,
1689                                                      $bad_attvals,
1690                                                      $add_attr_to_tag,
1691                                                      $message,
1692                                                      $id,
1693                                                      $mailbox
1694                                                      );
1695                             }
1696                         }
1697                     }
1698                 }
1699             }
1700             if ($tagname != false && $skip_content == false){
1701                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1702             }
1703         }
1704         $curpos = $gt+1;
1705     }
1706     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1707     if ($force_tag_closing == true){
1708         foreach ($open_tags as $tagname=>$opentimes){
1709             while ($opentimes > 0){
1710                 $trusted .= '</' . $tagname . '>';
1711                 $opentimes--;
1712             }
1713         }
1714         $trusted .= "\n";
1715     }
1716     $trusted .= "<!-- end sanitized html -->\n";
1717     return $trusted;
1718 }
1719
1720 /**
1721  * This is a wrapper function to call html sanitizing routines.
1722  *
1723  * @param  $body  the body of the message
1724  * @param  $id    the id of the message
1725  * @return        a string with html safe to display in the browser.
1726  */
1727 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1728     global $attachment_common_show_images, $view_unsafe_images,
1729            $has_unsafe_images;
1730     /**
1731      * Don't display attached images in HTML mode.
1732      */
1733     $attachment_common_show_images = false;
1734     $tag_list = Array(
1735                       false,
1736                       "object",
1737                       "meta",
1738                       "html",
1739                       "head",
1740                       "base",
1741                       "link",
1742                       "frame",
1743                       "iframe",
1744                       "plaintext",
1745                       "marquee"
1746                       );
1747
1748     $rm_tags_with_content = Array(
1749                                   "script",
1750                                   "applet",
1751                                   "embed",
1752                                   "title",
1753                                   "frameset",
1754                                   "xml"
1755                                   );
1756
1757     $self_closing_tags =  Array(
1758                                 "img",
1759                                 "br",
1760                                 "hr",
1761                                 "input"
1762                                 );
1763
1764     $force_tag_closing = true;
1765
1766     $rm_attnames = Array(
1767                          "/.*/" =>
1768                          Array(
1769                                "/target/i",
1770                                "/^on.*/i",
1771                                "/^dynsrc/i",
1772                                "/^data.*/i",
1773                                "/^lowsrc.*/i"
1774                                )
1775                          );
1776
1777     $secremoveimg = "../images/" . _("sec_remove_eng.png");
1778     $bad_attvals = Array(
1779         "/.*/" =>
1780             Array(
1781                 "/^src|background/i" =>
1782                     Array(
1783                           Array(
1784                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1785                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1786                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1787                                 ),
1788                           Array(
1789                                 "\\1$secremoveimg\\2",
1790                                 "\\1$secremoveimg\\2",
1791                                 "\\1$secremoveimg\\2",
1792                                 "\\1$secremoveimg\\2"
1793                                 )
1794                         ),
1795                 "/^href|action/i" =>
1796                     Array(
1797                           Array(
1798                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1799                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1800                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1801                                 ),
1802                           Array(
1803                                 "\\1#\\1",
1804                                 "\\1#\\1",
1805                                 "\\1#\\1",
1806                                 "\\1#\\1"
1807                                 )
1808                         ),
1809                 "/^style/i" =>
1810                     Array(
1811                           Array(
1812                                 "/expression/i",
1813                                 "/binding/i",
1814                                 "/behaviou*r/i",
1815                                 "/include-source/i",
1816                                 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
1817                                 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
1818                                 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si"
1819                                ),
1820                           Array(
1821                                 "idiocy",
1822                                 "idiocy",
1823                                 "idiocy",
1824                                 "idiocy",
1825                                 "url(\\1#\\1)",
1826                                 "url(\\1#\\1)",
1827                                 "url(\\1#\\1)",
1828                                 "url(\\1#\\1)"
1829                                )
1830                           )
1831                 )
1832         );
1833     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1834         $view_unsafe_images = false;
1835     }
1836     if (!$view_unsafe_images){
1837         /**
1838          * Remove any references to http/https if view_unsafe_images set
1839          * to false.
1840          */
1841          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1842                     '/^([\'\"])\s*https*:.*([\'\"])/si');
1843          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1844                     "\\1$secremoveimg\\1");
1845          array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
1846                     '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1847          array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
1848                     "url(\\1$secremoveimg\\1)");
1849     }
1850
1851     $add_attr_to_tag = Array(
1852         "/^a$/i" =>
1853             Array('target'=>'"_new"',
1854                   'title'=>'"'._("This external link will open in a new window").'"'
1855             )
1856     );
1857     $trusted = sq_sanitize($body,
1858                            $tag_list,
1859                            $rm_tags_with_content,
1860                            $self_closing_tags,
1861                            $force_tag_closing,
1862                            $rm_attnames,
1863                            $bad_attvals,
1864                            $add_attr_to_tag,
1865                            $message,
1866                            $id,
1867                            $mailbox
1868                            );
1869     if (preg_match("|$secremoveimg|i", $trusted)){
1870         $has_unsafe_images = true;
1871     }
1872     return $trusted;
1873 }
1874
1875 ?>