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 = $from_o->getAddress(false);
 445             } else {
 446                 $from_name = _("Unknown sender");
 447             }
 448             $from_name = decodeHeader(($from_name));
 449             $description = $from_name;
 450         } else {
 451             $default_page = SM_PATH . 'src/download.php';
 452             if (is_object($header->disposition)) {
 453                 $filename = $header->disposition->getProperty('filename');
 454                 if (trim($filename) == '') {
 455                     $name = decodeHeader($header->disposition->getProperty('name'));
 456                     if (trim($name) == '') {
 457                         $name = $header->getParameter('name');
 458                         if(trim($name) == '') {
 459                             if (trim( $header->id ) == '') {
 460                                 $filename = 'untitled-[' . $ent . ']' ;
 461                             } else {
 462                                 $filename = 'cid: ' . $header->id;
 463                             }
 464                         } else {
 465                             $filename = $name;
 466                         }
 467                     } else {
 468                         $filename = $name;
 469                     }
 470                 }
 471             } else {
 472                 $filename = $header->getParameter('name');
 473                 if (!trim($filename)) {
 474                     if (trim( $header->id ) == '') {
 475                         $filename = 'untitled-[' . $ent . ']' ;
 476                     } else {
 477                         $filename = 'cid: ' . $header->id;
 478                     }
 479                 }
 480             }
 481             if ($header->description) {
 482                 $description = decodeHeader($header->description);
 483             } else {
 484                 $description = '';
 485             }
 486         }
 487
 488         $display_filename = $filename;
 489         if (isset($passed_ent_id)) {
 490             $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
 491         } else {
 492             $passed_ent_id_link = '';
 493         }
 494         $defaultlink = $default_page . "?startMessage=$startMessage"
 495                      . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
 496                      . '&amp;ent_id='.$ent.$passed_ent_id_link;
 497         if ($where && $what) {
 498            $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
 499         }
 500         /* This executes the attachment hook with a specific MIME-type.
 501          * If that doesn't have results, it tries if there's a rule
 502          * for a more generic type.
 503          */
 504         $hookresults = do_hook("attachment $type0/$type1", $links,
 505                                $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 506                                $display_filename, $where, $what);
 507         if(count($hookresults[1]) <= 1) {
 508             $hookresults = do_hook("attachment $type0/*", $links,
 509                                    $startMessage, $id, $urlMailbox, $ent, $defaultlink,
 510                                    $display_filename, $where, $what);
 511         }
 512
 513         $links = $hookresults[1];
 514         $defaultlink = $hookresults[6];
 515
 516         $attachments .= '<TR><TD>' .
 517                         '<A HREF="'.$defaultlink.'">'.decodeHeader($display_filename).'</A>&nbsp;</TD>' .
 518                         '<TD><SMALL><b>' . show_readable_size($header->size) .
 519                         '</b>&nbsp;&nbsp;</small></TD>' .
 520                         "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
 521                         '<TD><SMALL>';
 522         $attachments .= '<b>' . $description . '</b>';
 523         $attachments .= '</SMALL></TD><TD><SMALL>&nbsp;';
 524
 525         $skipspaces = 1;
 526         foreach ($links as $val) {
 527             if ($skipspaces) {
 528                 $skipspaces = 0;
 529             } else {
 530                 $attachments .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
 531             }
 532             $attachments .= '<a href="' . $val['href'] . '">' .  $val['text'] . '</a>';
 533         }
 534         unset($links);
 535         $attachments .= "</TD></TR>\n";
 536     }
 537     return $attachments;
 538 }
 539
 540 function sqimap_base64_decode(&$string) {
 541     $string = str_replace("\r\n", "\n", $string);
 542     $string = base64_decode($string);
 543 }
 544
 545 /* This function decodes the body depending on the encoding type. */
 546 function decodeBody($body, $encoding) {
 547     global $show_html_default;
 548
 549     $body = str_replace("\r\n", "\n", $body);
 550     $encoding = strtolower($encoding);
 551
 552     if ($encoding == 'quoted-printable' ||
 553         $encoding == 'quoted_printable') {
 554         $body = quoted_printable_decode($body);
 555
 556         while (ereg("=\n", $body)) {
 557             $body = ereg_replace ("=\n", '', $body);
 558         }
 559
 560     } else if ($encoding == 'base64') {
 561         $body = base64_decode($body);
 562     }
 563
 564     // All other encodings are returned raw.
 565     return $body;
 566 }
 567
 568 /*
 569  * This functions decode strings that is encoded according to
 570  * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
 571  * Patched by Christian Schmidt <christian@ostenfeld.dk>  23/03/2002
 572  */
 573 function decodeHeader ($string, $utfencode=true,$htmlsave=true) {
 574     global $languages, $squirrelmail_language;
 575     if (is_array($string)) {
 576         $string = implode("\n", $string);
 577     }
 578
 579     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 580         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 581         $string = $languages[$squirrelmail_language]['XTRA_CODE']('decodeheader', $string);
 582         // Do we need to return at this point?
 583         // return $string;
 584     }
 585     $i = 0;
 586     $iLastMatch = -2;
 587     $encoded = false;
 588
 589     $aString = explode(' ',$string);
 590     $ret = '';
 591     foreach ($aString as $chunk) {
 592         if ($encoded && !$chunk) {
 593             continue;
 594         } elseif (!$chunk) {
 595             $ret .= ' ';
 596             continue;
 597         }
 598         $encoded = false;
 599         /* if encoded words are not separated by a linear-space-white we still catch them */
 600         $j = $i-1;
 601 //      if ($chunk{0} === '=') { /* performance, saves an unnessecarry preg call */
 602         while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
 603             /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
 604             if ($iLastMatch !== $j) {
 605                 if ($htmlsave) {
 606                     $ret .= '&nbsp;';
 607                 } else {
 608                     $ret .= ' ';
 609                 }
 610             }
 611             $iLastMatch = $i;
 612             $j = $i;
 613             $ret .= $res[1];
 614             $encoding = ucfirst($res[3]);
 615             switch ($encoding)
 616             {
 617             case 'B':
 618                 $replace = base64_decode($res[4]);
 619                 $ret .= charset_decode($res[2],$replace);
 620                 break;
 621             case 'Q':
 622                 $replace = str_replace('_', ' ', $res[4]);
 623                 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
 624                                     $replace);
 625                 /* Only encode into entities by default. Some places
 626                  * don't need the encoding, like the compose form.
 627                  */
 628                 if ($utfencode) {
 629                     $replace = charset_decode($res[2], $replace);
 630                 } else {
 631                     if ($htmlsave) {
 632                         $replace = htmlspecialchars($replace);
 633                     }
 634                 }
 635                 $ret .= $replace;
 636                 break;
 637             default:
 638                 break;
 639             }
 640             $chunk = $res[5];
 641             $encoded = true;
 642         }
 643 //      }
 644         if (!$encoded) {
 645             if ($htmlsave) {
 646                 $ret .= '&nbsp;';
 647             } else {
 648                 $ret .= ' ';
 649             }
 650         }
 651
 652         if (!$encoded && $htmlsave) {
 653             $ret .= htmlspecialchars($chunk);
 654         } else {
 655             $ret .= $chunk;
 656         }
 657         ++$i;
 658     }
 659     /* remove the first added space */
 660     if ($ret) {
 661         if ($htmlsave) {
 662             $ret = substr($ret,6);
 663         } else {
 664             $ret = substr($ret,1);
 665         }
 666     }
 667
 668     return $ret;
 669 }
 670
 671 /*
 672  * Encode a string according to RFC 1522 for use in headers if it
 673  * contains 8-bit characters or anything that looks like it should
 674  * be encoded.
 675  */
 676 function encodeHeader ($string) {
 677     global $default_charset, $languages, $squirrelmail_language;
 678
 679     if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
 680         function_exists($languages[$squirrelmail_language]['XTRA_CODE'])) {
 681         return  $languages[$squirrelmail_language]['XTRA_CODE']('encodeheader', $string);
 682     }
 683     if (strtolower($default_charset) == 'iso-8859-1') {
 684         $string = str_replace("\240",' ',$string);
 685     }
 686
 687     // Encode only if the string contains 8-bit characters or =?
 688     $j = strlen($string);
 689     $max_l = 75 - strlen($default_charset) - 7;
 690     $aRet = array();
 691     $ret = '';
 692     $iEncStart = $enc_init = false;
 693     $cur_l = $iOffset = 0;
 694     for($i = 0; $i < $j; ++$i) {
 695         switch($string{$i})
 696         {
 697         case '=':
 698         case '<':
 699         case '>':
 700         case ',':
 701         case '?':
 702         case '_':
 703             if ($iEncStart === false) {
 704                 $iEncStart = $i;
 705             }
 706             $cur_l+=3;
 707             if ($cur_l > ($max_l-2)) {
 708                 /* if there is an stringpart that doesn't need encoding, add it */
 709                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 710                 $aRet[] = "=?$default_charset?Q?$ret?=";
 711                 $iOffset = $i;
 712                 $cur_l = 0;
 713                 $ret = '';
 714                 $iEncStart = false;
 715             } else {
 716                 $ret .= sprintf("=%02X",ord($string{$i}));
 717             }
 718             break;
 719         case '(':
 720         case ')':
 721             if ($iEncStart !== false) {
 722                 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 723                 $aRet[] = "=?$default_charset?Q?$ret?=";
 724                 $iOffset = $i;
 725                 $cur_l = 0;
 726                 $ret = '';
 727                 $iEncStart = false;
 728             }
 729             break;
 730         case ' ':
 731             if ($iEncStart !== false) {
 732                 $cur_l++;
 733                 if ($cur_l > $max_l) {
 734                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 735                     $aRet[] = "=?$default_charset?Q?$ret?=";
 736                     $iOffset = $i;
 737                     $cur_l = 0;
 738                     $ret = '';
 739                     $iEncStart = false;
 740                 } else {
 741                     $ret .= '_';
 742                 }
 743             }
 744             break;
 745         default:
 746             $k = ord($string{$i});
 747             if ($k > 126) {
 748                 if ($iEncStart === false) {
 749                     // do not start encoding in the middle of a string, also take the rest of the word.
 750                     $sLeadString = substr($string,0,$i);
 751                     $aLeadString = explode(' ',$sLeadString);
 752                     $sToBeEncoded = array_pop($aLeadString);
 753                     $iEncStart = $i - strlen($sToBeEncoded);
 754                     $ret .= $sToBeEncoded;
 755                     $cur_l += strlen($sToBeEncoded);
 756                 }
 757                 $cur_l += 3;
 758                 /* first we add the encoded string that reached it's max size */
 759                 if ($cur_l > ($max_l-2)) {
 760                     $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 761                     $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
 762                     $cur_l = 3;
 763                     $ret = '';
 764                     $iOffset = $i;
 765                     $iEncStart = $i;
 766                 }
 767                 $enc_init = true;
 768                 $ret .= sprintf("=%02X", $k);
 769             } else {
 770                 if ($iEncStart !== false) {
 771                     $cur_l++;
 772                     if ($cur_l > $max_l) {
 773                         $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 774                         $aRet[] = "=?$default_charset?Q?$ret?=";
 775                         $iEncStart = false;
 776                         $iOffset = $i;
 777                         $cur_l = 0;
 778                         $ret = '';
 779                     } else {
 780                         $ret .= $string{$i};
 781                     }
 782                 }
 783             }
 784             break;
 785         }
 786     }
 787
 788     if ($enc_init) {
 789         if ($iEncStart !== false) {
 790             $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
 791             $aRet[] = "=?$default_charset?Q?$ret?=";
 792         } else {
 793             $aRet[] = substr($string,$iOffset);
 794         }
 795         $string = implode('',$aRet);
 796     }
 797     return $string;
 798 }
 799
 800 /* This function trys to locate the entity_id of a specific mime element */
 801 function find_ent_id($id, $message) {
 802     for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
 803         if ($message->entities[$i]->header->type0 == 'multipart')  {
 804             $ret = find_ent_id($id, $message->entities[$i]);
 805         } else {
 806             if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
 807 //                if (sq_check_save_extension($message->entities[$i])) {
 808                     return $message->entities[$i]->entity_id;
 809 //                }
 810             }
 811         }
 812     }
 813     return $ret;
 814 }
 815
 816 function sq_check_save_extension($message) {
 817     $filename = $message->getFilename();
 818     $ext = substr($filename, strrpos($filename,'.')+1);
 819     $save_extensions = array('jpg','jpeg','gif','png','bmp');
 820     return in_array($ext, $save_extensions);
 821 }
 822
 823
 824 /**
 825  ** HTMLFILTER ROUTINES
 826  */
 827
 828 /**
 829  * This function returns the final tag out of the tag name, an array
 830  * of attributes, and the type of the tag. This function is called by
 831  * sq_sanitize internally.
 832  *
 833  * @param  $tagname  the name of the tag.
 834  * @param  $attary   the array of attributes and their values
 835  * @param  $tagtype  The type of the tag (see in comments).
 836  * @return           a string with the final tag representation.
 837  */
 838 function sq_tagprint($tagname, $attary, $tagtype){
 839     $me = 'sq_tagprint';
 840
 841     if ($tagtype == 2){
 842         $fulltag = '</' . $tagname . '>';
 843     } else {
 844         $fulltag = '<' . $tagname;
 845         if (is_array($attary) && sizeof($attary)){
 846             $atts = Array();
 847             while (list($attname, $attvalue) = each($attary)){
 848                 array_push($atts, "$attname=$attvalue");
 849             }
 850             $fulltag .= ' ' . join(" ", $atts);
 851         }
 852         if ($tagtype == 3){
 853             $fulltag .= ' /';
 854         }
 855         $fulltag .= '>';
 856     }
 857     return $fulltag;
 858 }
 859
 860 /**
 861  * A small helper function to use with array_walk. Modifies a by-ref
 862  * value and makes it lowercase.
 863  *
 864  * @param  $val a value passed by-ref.
 865  * @return      void since it modifies a by-ref value.
 866  */
 867 function sq_casenormalize(&$val){
 868     $val = strtolower($val);
 869 }
 870
 871 /**
 872  * This function skips any whitespace from the current position within
 873  * a string and to the next non-whitespace value.
 874  *
 875  * @param  $body   the string
 876  * @param  $offset the offset within the string where we should start
 877  *                 looking for the next non-whitespace character.
 878  * @return         the location within the $body where the next
 879  *                 non-whitespace char is located.
 880  */
 881 function sq_skipspace($body, $offset){
 882     $me = 'sq_skipspace';
 883     preg_match('/^(\s*)/s', substr($body, $offset), $matches);
 884     if (sizeof($matches{1})){
 885         $count = strlen($matches{1});
 886         $offset += $count;
 887     }
 888     return $offset;
 889 }
 890
 891 /**
 892  * This function looks for the next character within a string.  It's
 893  * really just a glorified "strpos", except it catches if failures
 894  * nicely.
 895  *
 896  * @param  $body   The string to look for needle in.
 897  * @param  $offset Start looking from this position.
 898  * @param  $needle The character/string to look for.
 899  * @return         location of the next occurance of the needle, or
 900  *                 strlen($body) if needle wasn't found.
 901  */
 902 function sq_findnxstr($body, $offset, $needle){
 903     $me  = 'sq_findnxstr';
 904     $pos = strpos($body, $needle, $offset);
 905     if ($pos === FALSE){
 906         $pos = strlen($body);
 907     }
 908     return $pos;
 909 }
 910
 911 /**
 912  * This function takes a PCRE-style regexp and tries to match it
 913  * within the string.
 914  *
 915  * @param  $body   The string to look for needle in.
 916  * @param  $offset Start looking from here.
 917  * @param  $reg    A PCRE-style regex to match.
 918  * @return         Returns a false if no matches found, or an array
 919  *                 with the following members:
 920  *                 - integer with the location of the match within $body
 921  *                 - string with whatever content between offset and the match
 922  *                 - string with whatever it is we matched
 923  */
 924 function sq_findnxreg($body, $offset, $reg){
 925     $me = 'sq_findnxreg';
 926     $matches = Array();
 927     $retarr = Array();
 928     preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
 929     if (!isset($matches{0}) || !$matches{0}){
 930         $retarr = false;
 931     } else {
 932         $retarr{0} = $offset + strlen($matches{1});
 933         $retarr{1} = $matches{1};
 934         $retarr{2} = $matches{2};
 935     }
 936     return $retarr;
 937 }
 938
 939 /**
 940  * This function looks for the next tag.
 941  *
 942  * @param  $body   String where to look for the next tag.
 943  * @param  $offset Start looking from here.
 944  * @return         false if no more tags exist in the body, or
 945  *                 an array with the following members:
 946  *                 - string with the name of the tag
 947  *                 - array with attributes and their values
 948  *                 - integer with tag type (1, 2, or 3)
 949  *                 - integer where the tag starts (starting "<")
 950  *                 - integer where the tag ends (ending ">")
 951  *                 first three members will be false, if the tag is invalid.
 952  */
 953 function sq_getnxtag($body, $offset){
 954     $me = 'sq_getnxtag';
 955     if ($offset > strlen($body)){
 956         return false;
 957     }
 958     $lt = sq_findnxstr($body, $offset, "<");
 959     if ($lt == strlen($body)){
 960         return false;
 961     }
 962     /**
 963      * We are here:
 964      * blah blah <tag attribute="value">
 965      * \---------^
 966      */
 967     $pos = sq_skipspace($body, $lt+1);
 968     if ($pos >= strlen($body)){
 969         return Array(false, false, false, $lt, strlen($body));
 970     }
 971     /**
 972      * There are 3 kinds of tags:
 973      * 1. Opening tag, e.g.:
 974      *    <a href="blah">
 975      * 2. Closing tag, e.g.:
 976      *    </a>
 977      * 3. XHTML-style content-less tag, e.g.:
 978      *    <img src="blah"/>
 979      */
 980     $tagtype = false;
 981     switch (substr($body, $pos, 1)){
 982         case '/':
 983             $tagtype = 2;
 984             $pos++;
 985             break;
 986         case '!':
 987             /**
 988              * A comment or an SGML declaration.
 989              */
 990             if (substr($body, $pos+1, 2) == "--"){
 991                 $gt = strpos($body, "-->", $pos);
 992                 if ($gt === false){
 993                     $gt = strlen($body);
 994                 } else {
 995                     $gt += 2;
 996                 }
 997                 return Array(false, false, false, $lt, $gt);
 998             } else {
 999                 $gt = sq_findnxstr($body, $pos, ">");
1000                 return Array(false, false, false, $lt, $gt);
1001             }
1002             break;
1003         default:
1004             /**
1005              * Assume tagtype 1 for now. If it's type 3, we'll switch values
1006              * later.
1007              */
1008             $tagtype = 1;
1009             break;
1010     }
1011
1012     $tag_start = $pos;
1013     $tagname = '';
1014     /**
1015      * Look for next [\W-_], which will indicate the end of the tag name.
1016      */
1017     $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1018     if ($regary == false){
1019         return Array(false, false, false, $lt, strlen($body));
1020     }
1021     list($pos, $tagname, $match) = $regary;
1022     $tagname = strtolower($tagname);
1023
1024     /**
1025      * $match can be either of these:
1026      * '>'  indicating the end of the tag entirely.
1027      * '\s' indicating the end of the tag name.
1028      * '/'  indicating that this is type-3 xhtml tag.
1029      *
1030      * Whatever else we find there indicates an invalid tag.
1031      */
1032     switch ($match){
1033         case '/':
1034             /**
1035              * This is an xhtml-style tag with a closing / at the
1036              * end, like so: <img src="blah"/>. Check if it's followed
1037              * by the closing bracket. If not, then this tag is invalid
1038              */
1039             if (substr($body, $pos, 2) == "/>"){
1040                 $pos++;
1041                 $tagtype = 3;
1042             } else {
1043                 $gt = sq_findnxstr($body, $pos, ">");
1044                 $retary = Array(false, false, false, $lt, $gt);
1045                 return $retary;
1046             }
1047         case '>':
1048             return Array($tagname, false, $tagtype, $lt, $pos);
1049             break;
1050         default:
1051             /**
1052              * Check if it's whitespace
1053              */
1054             if (!preg_match('/\s/', $match)){
1055                 /**
1056                  * This is an invalid tag! Look for the next closing ">".
1057                  */
1058                 $gt = sq_findnxstr($body, $lt, ">");
1059                 return Array(false, false, false, $lt, $gt);
1060             }
1061             break;
1062     }
1063
1064     /**
1065      * At this point we're here:
1066      * <tagname  attribute='blah'>
1067      * \-------^
1068      *
1069      * At this point we loop in order to find all attributes.
1070      */
1071     $attname = '';
1072     $atttype = false;
1073     $attary = Array();
1074
1075     while ($pos <= strlen($body)){
1076         $pos = sq_skipspace($body, $pos);
1077         if ($pos == strlen($body)){
1078             /**
1079              * Non-closed tag.
1080              */
1081             return Array(false, false, false, $lt, $pos);
1082         }
1083         /**
1084          * See if we arrived at a ">" or "/>", which means that we reached
1085          * the end of the tag.
1086          */
1087         $matches = Array();
1088         if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1089             /**
1090              * Yep. So we did.
1091              */
1092             $pos += strlen($matches{1});
1093             if ($matches{2} == "/>"){
1094                 $tagtype = 3;
1095                 $pos++;
1096             }
1097             return Array($tagname, $attary, $tagtype, $lt, $pos);
1098         }
1099
1100         /**
1101          * There are several types of attributes, with optional
1102          * [:space:] between members.
1103          * Type 1:
1104          *   attrname[:space:]=[:space:]'CDATA'
1105          * Type 2:
1106          *   attrname[:space:]=[:space:]"CDATA"
1107          * Type 3:
1108          *   attr[:space:]=[:space:]CDATA
1109          * Type 4:
1110          *   attrname
1111          *
1112          * We leave types 1 and 2 the same, type 3 we check for
1113          * '"' and convert to "&quot" if needed, then wrap in
1114          * double quotes. Type 4 we convert into:
1115          * attrname="yes".
1116          */
1117         $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1118         if ($regary == false){
1119             /**
1120              * Looks like body ended before the end of tag.
1121              */
1122             return Array(false, false, false, $lt, strlen($body));
1123         }
1124         list($pos, $attname, $match) = $regary;
1125         $attname = strtolower($attname);
1126         /**
1127          * We arrived at the end of attribute name. Several things possible
1128          * here:
1129          * '>'  means the end of the tag and this is attribute type 4
1130          * '/'  if followed by '>' means the same thing as above
1131          * '\s' means a lot of things -- look what it's followed by.
1132          *      anything else means the attribute is invalid.
1133          */
1134         switch($match){
1135             case '/':
1136                 /**
1137                  * This is an xhtml-style tag with a closing / at the
1138                  * end, like so: <img src="blah"/>. Check if it's followed
1139                  * by the closing bracket. If not, then this tag is invalid
1140                  */
1141                 if (substr($body, $pos, 2) == "/>"){
1142                     $pos++;
1143                     $tagtype = 3;
1144                 } else {
1145                     $gt = sq_findnxstr($body, $pos, ">");
1146                     $retary = Array(false, false, false, $lt, $gt);
1147                     return $retary;
1148                 }
1149             case '>':
1150                 $attary{$attname} = '"yes"';
1151                 return Array($tagname, $attary, $tagtype, $lt, $pos);
1152                 break;
1153             default:
1154                 /**
1155                  * Skip whitespace and see what we arrive at.
1156                  */
1157                 $pos = sq_skipspace($body, $pos);
1158                 $char = substr($body, $pos, 1);
1159                 /**
1160                  * Two things are valid here:
1161                  * '=' means this is attribute type 1 2 or 3.
1162                  * \w means this was attribute type 4.
1163                  * anything else we ignore and re-loop. End of tag and
1164                  * invalid stuff will be caught by our checks at the beginning
1165                  * of the loop.
1166                  */
1167                 if ($char == "="){
1168                     $pos++;
1169                     $pos = sq_skipspace($body, $pos);
1170                     /**
1171                      * Here are 3 possibilities:
1172                      * "'"  attribute type 1
1173                      * '"'  attribute type 2
1174                      * everything else is the content of tag type 3
1175                      */
1176                     $quot = substr($body, $pos, 1);
1177                     if ($quot == "'"){
1178                         $regary = sq_findnxreg($body, $pos+1, "\'");
1179                         if ($regary == false){
1180                             return Array(false, false, false, $lt, strlen($body));
1181                         }
1182                         list($pos, $attval, $match) = $regary;
1183                         $pos++;
1184                         $attary{$attname} = "'" . $attval . "'";
1185                     } else if ($quot == '"'){
1186                         $regary = sq_findnxreg($body, $pos+1, '\"');
1187                         if ($regary == false){
1188                             return Array(false, false, false, $lt, strlen($body));
1189                         }
1190                         list($pos, $attval, $match) = $regary;
1191                         $pos++;
1192                         $attary{$attname} = '"' . $attval . '"';
1193                     } else {
1194                         /**
1195                          * These are hateful. Look for \s, or >.
1196                          */
1197                         $regary = sq_findnxreg($body, $pos, "[\s>]");
1198                         if ($regary == false){
1199                             return Array(false, false, false, $lt, strlen($body));
1200                         }
1201                         list($pos, $attval, $match) = $regary;
1202                         /**
1203                          * If it's ">" it will be caught at the top.
1204                          */
1205                         $attval = preg_replace("/\"/s", "&quot;", $attval);
1206                         $attary{$attname} = '"' . $attval . '"';
1207                     }
1208                 } else if (preg_match("|[\w/>]|", $char)) {
1209                     /**
1210                      * That was attribute type 4.
1211                      */
1212                     $attary{$attname} = '"yes"';
1213                 } else {
1214                     /**
1215                      * An illegal character. Find next '>' and return.
1216                      */
1217                     $gt = sq_findnxstr($body, $pos, ">");
1218                     return Array(false, false, false, $lt, $gt);
1219                 }
1220                 break;
1221         }
1222     }
1223     /**
1224      * The fact that we got here indicates that the tag end was never
1225      * found. Return invalid tag indication so it gets stripped.
1226      */
1227     return Array(false, false, false, $lt, strlen($body));
1228 }
1229
1230 /**
1231  * This function checks attribute values for entity-encoded values
1232  * and returns them translated into 8-bit strings so we can run
1233  * checks on them.
1234  *
1235  * @param  $attvalue A string to run entity check against.
1236  * @return           Translated value.
1237  */
1238 function sq_deent($attvalue){
1239     $me = 'sq_deent';
1240     /**
1241      * See if we have to run the checks first. All entities must start
1242      * with "&".
1243      */
1244     if (strpos($attvalue, "&") === false){
1245         return $attvalue;
1246     }
1247     /**
1248      * Check named entities first.
1249      */
1250     $trans = get_html_translation_table(HTML_ENTITIES);
1251     /**
1252      * Leave &quot; in, as it can mess us up.
1253      */
1254     $trans = array_flip($trans);
1255     unset($trans{"&quot;"});
1256     while (list($ent, $val) = each($trans)){
1257         $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1258     }
1259     /**
1260      * Now translate numbered entities from 1 to 255 if needed.
1261      */
1262     if (strpos($attvalue, "#") !== false){
1263         $omit = Array(34, 39);
1264         for ($asc=1; $asc<256; $asc++){
1265             if (!in_array($asc, $omit)){
1266                 $chr = chr($asc);
1267                 $attvalue = preg_replace("/\&#0*$asc;*(\D)/si", "$chr\\1",
1268                                          $attvalue);
1269                 $attvalue = preg_replace("/\&#x0*".dechex($asc).";*(\W)/si",
1270                                          "$chr\\1", $attvalue);
1271             }
1272         }
1273     }
1274     return $attvalue;
1275 }
1276
1277 /**
1278  * This function runs various checks against the attributes.
1279  *
1280  * @param  $tagname         String with the name of the tag.
1281  * @param  $attary          Array with all tag attributes.
1282  * @param  $rm_attnames     See description for sq_sanitize
1283  * @param  $bad_attvals     See description for sq_sanitize
1284  * @param  $add_attr_to_tag See description for sq_sanitize
1285  * @param  $message         message object
1286  * @param  $id              message id
1287  * @return                  Array with modified attributes.
1288  */
1289 function sq_fixatts($tagname,
1290                     $attary,
1291                     $rm_attnames,
1292                     $bad_attvals,
1293                     $add_attr_to_tag,
1294                     $message,
1295                     $id,
1296                     $mailbox
1297                     ){
1298     $me = 'sq_fixatts';
1299     while (list($attname, $attvalue) = each($attary)){
1300         /**
1301          * See if this attribute should be removed.
1302          */
1303         foreach ($rm_attnames as $matchtag=>$matchattrs){
1304             if (preg_match($matchtag, $tagname)){
1305                 foreach ($matchattrs as $matchattr){
1306                     if (preg_match($matchattr, $attname)){
1307                         unset($attary{$attname});
1308                         continue;
1309                     }
1310                 }
1311             }
1312         }
1313         /**
1314          * Remove any entities.
1315          */
1316         $attvalue = sq_deent($attvalue);
1317
1318         /**
1319          * Now let's run checks on the attvalues.
1320          * I don't expect anyone to comprehend this. If you do,
1321          * get in touch with me so I can drive to where you live and
1322          * shake your hand personally. :)
1323          */
1324         foreach ($bad_attvals as $matchtag=>$matchattrs){
1325             if (preg_match($matchtag, $tagname)){
1326                 foreach ($matchattrs as $matchattr=>$valary){
1327                     if (preg_match($matchattr, $attname)){
1328                         /**
1329                          * There are two arrays in valary.
1330                          * First is matches.
1331                          * Second one is replacements
1332                          */
1333                         list($valmatch, $valrepl) = $valary;
1334                         $newvalue =
1335                             preg_replace($valmatch, $valrepl, $attvalue);
1336                         if ($newvalue != $attvalue){
1337                             $attary{$attname} = $newvalue;
1338                         }
1339                     }
1340                 }
1341             }
1342         }
1343         /**
1344          * Turn cid: urls into http-friendly ones.
1345          */
1346         if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1347             $attary{$attname} = sq_cid2http($message, $id, $attvalue, $mailbox);
1348         }
1349     }
1350     /**
1351      * See if we need to append any attributes to this tag.
1352      */
1353     foreach ($add_attr_to_tag as $matchtag=>$addattary){
1354         if (preg_match($matchtag, $tagname)){
1355             $attary = array_merge($attary, $addattary);
1356         }
1357     }
1358     return $attary;
1359 }
1360
1361 /**
1362  * This function edits the style definition to make them friendly and
1363  * usable in squirrelmail.
1364  *
1365  * @param  $message  the message object
1366  * @param  $id       the message id
1367  * @param  $content  a string with whatever is between <style> and </style>
1368  * @return           a string with edited content.
1369  */
1370 function sq_fixstyle($body, $pos, $message, $id){
1371     global $view_unsafe_images;
1372     $me = 'sq_fixstyle';
1373     $ret = sq_findnxreg($body, $pos, '</\s*style\s*>');
1374     if ($ret == FALSE){
1375         return array(FALSE, strlen($body));
1376     }
1377     $newpos = $ret[0] + strlen($ret[2]);
1378     $content = $ret[1];
1379     /**
1380      * First look for general BODY style declaration, which would be
1381      * like so:
1382      * body {background: blah-blah}
1383      * and change it to .bodyclass so we can just assign it to a <div>
1384      */
1385     $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1386     $secremoveimg = '../images/' . _("sec_remove_eng.png");
1387     /**
1388      * Fix url('blah') declarations.
1389      */
1390     $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1391                             "url(\\1$secremoveimg\\2)", $content);
1392     /**
1393      * Fix url('https*://.*) declarations but only if $view_unsafe_images
1394      * is false.
1395      */
1396     if (!$view_unsafe_images){
1397         $content = preg_replace("|url\s*\(\s*([\'\"])\s*https*:.*?([\'\"])\s*\)|si",
1398                                 "url(\\1$secremoveimg\\2)", $content);
1399     }
1400
1401     /**
1402      * Fix urls that refer to cid:
1403      */
1404     while (preg_match("|url\s*\(\s*([\'\"]\s*cid:.*?[\'\"])\s*\)|si",
1405                       $content, $matches)){
1406         $cidurl = $matches{1};
1407         $httpurl = sq_cid2http($message, $id, $cidurl);
1408         $content = preg_replace("|url\s*\(\s*$cidurl\s*\)|si",
1409                                 "url($httpurl)", $content);
1410     }
1411
1412     /**
1413      * Fix stupid css declarations which lead to vulnerabilities
1414      * in IE.
1415      */
1416     $match   = Array('/expression/i',
1417                      '/behaviou*r/i',
1418                      '/binding/i');
1419     $replace = Array('idiocy', 'idiocy', 'idiocy');
1420     $content = preg_replace($match, $replace, $content);
1421     return array($content, $newpos);
1422 }
1423
1424 /**
1425  * This function converts cid: url's into the ones that can be viewed in
1426  * the browser.
1427  *
1428  * @param  $message  the message object
1429  * @param  $id       the message id
1430  * @param  $cidurl   the cid: url.
1431  * @return           a string with a http-friendly url
1432  */
1433 function sq_cid2http($message, $id, $cidurl, $mailbox){
1434     /**
1435      * Get rid of quotes.
1436      */
1437     $quotchar = substr($cidurl, 0, 1);
1438     $cidurl = str_replace($quotchar, "", $cidurl);
1439     $cidurl = substr(trim($cidurl), 4);
1440     $linkurl = find_ent_id($cidurl, $message);
1441     /* in case of non-save cid links $httpurl should be replaced by a sort of
1442        unsave link image */
1443     $httpurl = '';
1444     if ($linkurl) {
1445         $httpurl = $quotchar . SM_PATH . 'src/download.php?absolute_dl=true&amp;' .
1446                    "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
1447                    '&amp;ent_id=' . $linkurl . $quotchar;
1448     }
1449     return $httpurl;
1450 }
1451
1452 /**
1453  * This function changes the <body> tag into a <div> tag since we
1454  * can't really have a body-within-body.
1455  *
1456  * @param  $attary  an array of attributes and values of <body>
1457  * @return          a modified array of attributes to be set for <div>
1458  */
1459 function sq_body2div($attary){
1460     $me = 'sq_body2div';
1461     $divattary = Array('class' => "'bodyclass'");
1462     $bgcolor = '#ffffff';
1463     $text = '#000000';
1464     $styledef = '';
1465     if (is_array($attary) && sizeof($attary) > 0){
1466         foreach ($attary as $attname=>$attvalue){
1467             $quotchar = substr($attvalue, 0, 1);
1468             $attvalue = str_replace($quotchar, "", $attvalue);
1469             switch ($attname){
1470                 case 'background':
1471                     $styledef .= "background-image: url('$attvalue'); ";
1472                     break;
1473                 case 'bgcolor':
1474                     $styledef .= "background-color: $attvalue; ";
1475                     break;
1476                 case 'text':
1477                     $styledef .= "color: $attvalue; ";
1478                     break;
1479             }
1480         }
1481         if (strlen($styledef) > 0){
1482             $divattary{"style"} = "\"$styledef\"";
1483         }
1484     }
1485     return $divattary;
1486 }
1487
1488 /**
1489  * This is the main function and the one you should actually be calling.
1490  * There are several variables you should be aware of an which need
1491  * special description.
1492  *
1493  * Since the description is quite lengthy, see it here:
1494  * http://www.mricon.com/html/phpfilter.html
1495  *
1496  * @param $body                 the string with HTML you wish to filter
1497  * @param $tag_list             see description above
1498  * @param $rm_tags_with_content see description above
1499  * @param $self_closing_tags    see description above
1500  * @param $force_tag_closing    see description above
1501  * @param $rm_attnames          see description above
1502  * @param $bad_attvals          see description above
1503  * @param $add_attr_to_tag      see description above
1504  * @param $message              message object
1505  * @param $id                   message id
1506  * @return                      sanitized html safe to show on your pages.
1507  */
1508 function sq_sanitize($body,
1509                      $tag_list,
1510                      $rm_tags_with_content,
1511                      $self_closing_tags,
1512                      $force_tag_closing,
1513                      $rm_attnames,
1514                      $bad_attvals,
1515                      $add_attr_to_tag,
1516                      $message,
1517                      $id,
1518                      $mailbox
1519                      ){
1520     $me = 'sq_sanitize';
1521     $rm_tags = array_shift($tag_list);
1522     /**
1523      * Normalize rm_tags and rm_tags_with_content.
1524      */
1525     @array_walk($tag_list, 'sq_casenormalize');
1526     @array_walk($rm_tags_with_content, 'sq_casenormalize');
1527     @array_walk($self_closing_tags, 'sq_casenormalize');
1528     /**
1529      * See if tag_list is of tags to remove or tags to allow.
1530      * false  means remove these tags
1531      * true   means allow these tags
1532      */
1533     $curpos = 0;
1534     $open_tags = Array();
1535     $trusted = "<!-- begin sanitized html -->\n";
1536     $skip_content = false;
1537     /**
1538      * Take care of netscape's stupid javascript entities like
1539      * &{alert('boo')};
1540      */
1541     $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
1542
1543     while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
1544         list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1545         $free_content = substr($body, $curpos, $lt-$curpos);
1546         /**
1547          * Take care of <style>
1548          */
1549         if ($tagname == "style" && $tagtype == 1){
1550             list($free_content, $curpos) =
1551                 sq_fixstyle($body, $gt+1, $message, $id);
1552             if ($free_content != FALSE){
1553                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1554                 $trusted .= $free_content;
1555                 $trusted .= sq_tagprint($tagname, false, 2);
1556             }
1557             continue;
1558         }
1559         if ($skip_content == false){
1560             $trusted .= $free_content;
1561         }
1562         if ($tagname != FALSE){
1563             if ($tagtype == 2){
1564                 if ($skip_content == $tagname){
1565                     /**
1566                      * Got to the end of tag we needed to remove.
1567                      */
1568                     $tagname = false;
1569                     $skip_content = false;
1570                 } else {
1571                     if ($skip_content == false){
1572                         if ($tagname == "body"){
1573                             $tagname = "div";
1574                         } else {
1575                             if (isset($open_tags{$tagname}) &&
1576                                 $open_tags{$tagname} > 0){
1577                                 $open_tags{$tagname}--;
1578                             } else {
1579                                 $tagname = false;
1580                             }
1581                         }
1582                     }
1583                 }
1584             } else {
1585                 /**
1586                  * $rm_tags_with_content
1587                  */
1588                 if ($skip_content == false){
1589                     /**
1590                      * See if this is a self-closing type and change
1591                      * tagtype appropriately.
1592                      */
1593                     if ($tagtype == 1
1594                         && in_array($tagname, $self_closing_tags)){
1595                         $tagtype=3;
1596                     }
1597                     /**
1598                      * See if we should skip this tag and any content
1599                      * inside it.
1600                      */
1601                     if ($tagtype == 1 &&
1602                         in_array($tagname, $rm_tags_with_content)){
1603                         $skip_content = $tagname;
1604                     } else {
1605                         if (($rm_tags == false
1606                              && in_array($tagname, $tag_list)) ||
1607                             ($rm_tags == true &&
1608                              !in_array($tagname, $tag_list))){
1609                             $tagname = false;
1610                         } else {
1611                             if ($tagtype == 1){
1612                                 if (isset($open_tags{$tagname})){
1613                                     $open_tags{$tagname}++;
1614                                 } else {
1615                                     $open_tags{$tagname}=1;
1616                                 }
1617                             }
1618                             /**
1619                              * This is where we run other checks.
1620                              */
1621                             if (is_array($attary) && sizeof($attary) > 0){
1622                                 $attary = sq_fixatts($tagname,
1623                                                      $attary,
1624                                                      $rm_attnames,
1625                                                      $bad_attvals,
1626                                                      $add_attr_to_tag,
1627                                                      $message,
1628                                                      $id,
1629                                                      $mailbox
1630                                                      );
1631                             }
1632                             /**
1633                              * Convert body into div.
1634                              */
1635                             if ($tagname == "body"){
1636                                 $tagname = "div";
1637                                 $attary = sq_body2div($attary, $message, $id);
1638                             }
1639                         }
1640                     }
1641                 }
1642             }
1643             if ($tagname != false && $skip_content == false){
1644                 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1645             }
1646         }
1647         $curpos = $gt+1;
1648     }
1649     $trusted .= substr($body, $curpos, strlen($body)-$curpos);
1650     if ($force_tag_closing == true){
1651         foreach ($open_tags as $tagname=>$opentimes){
1652             while ($opentimes > 0){
1653                 $trusted .= '</' . $tagname . '>';
1654                 $opentimes--;
1655             }
1656         }
1657         $trusted .= "\n";
1658     }
1659     $trusted .= "<!-- end sanitized html -->\n";
1660     return $trusted;
1661 }
1662
1663 /**
1664  * This is a wrapper function to call html sanitizing routines.
1665  *
1666  * @param  $body  the body of the message
1667  * @param  $id    the id of the message
1668  * @return        a string with html safe to display in the browser.
1669  */
1670 function magicHTML($body, $id, $message, $mailbox = 'INBOX') {
1671     global $attachment_common_show_images, $view_unsafe_images,
1672            $has_unsafe_images;
1673     /**
1674      * Don't display attached images in HTML mode.
1675      */
1676     $attachment_common_show_images = false;
1677     $tag_list = Array(
1678                       false,
1679                       "object",
1680                       "meta",
1681                       "html",
1682                       "head",
1683                       "base",
1684                       "link",
1685                       "frame",
1686                       "iframe"
1687                       );
1688
1689     $rm_tags_with_content = Array(
1690                                   "script",
1691                                   "applet",
1692                                   "embed",
1693                                   "title"
1694                                   );
1695
1696     $self_closing_tags =  Array(
1697                                 "img",
1698                                 "br",
1699                                 "hr",
1700                                 "input"
1701                                 );
1702
1703     $force_tag_closing = false;
1704
1705     $rm_attnames = Array(
1706                          "/.*/" =>
1707                          Array(
1708                                "/target/i",
1709                                "/^on.*/i",
1710                                "/^dynsrc/i",
1711                                "/^data.*/i",
1712                                "/^lowsrc.*/i"
1713                                )
1714                          );
1715
1716     $secremoveimg = "../images/" . _("sec_remove_eng.png");
1717     $bad_attvals = Array(
1718         "/.*/" =>
1719             Array(
1720                 "/^src|background/i" =>
1721                     Array(
1722                           Array(
1723                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1724                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1725                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1726                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1727                                 ),
1728                           Array(
1729                                 "\\1$secremoveimg\\2",
1730                                 "\\1$secremoveimg\\2",
1731                                 "\\1$secremoveimg\\2",
1732                                 "\\1$secremoveimg\\2"
1733                                 )
1734                         ),
1735                 "/^href|action/i" =>
1736                     Array(
1737                           Array(
1738                                 "|^([\'\"])\s*\.\./.*([\'\"])|si",
1739                                 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
1740                                 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
1741                                 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
1742                                 ),
1743                           Array(
1744                                 "\\1#\\2",
1745                                 "\\1#\\2",
1746                                 "\\1#\\2",
1747                                 "\\1#\\2"
1748                                 )
1749                         ),
1750                 "/^style/i" =>
1751                     Array(
1752                           Array(
1753                                 "/expression/i",
1754                                 "/binding/i",
1755                                 "/behaviou*r/i",
1756                                 "|url\s*\(\s*([\'\"])\s*\.\./.*([\'\"])\s*\)|si",
1757                                 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
1758                                 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
1759                                 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si"
1760                                ),
1761                           Array(
1762                                 "idiocy",
1763                                 "idiocy",
1764                                 "idiocy",
1765                                 "url(\\1#\\2)",
1766                                 "url(\\1#\\2)",
1767                                 "url(\\1#\\2)",
1768                                 "url(\\1#\\2)"
1769                                )
1770                           )
1771                 )
1772         );
1773     if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET) ) {
1774         $view_unsafe_images = false;
1775     }
1776     if (!$view_unsafe_images){
1777         /**
1778          * Remove any references to http/https if view_unsafe_images set
1779          * to false.
1780          */
1781          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
1782                     '/^([\'\"])\s*https*:.*([\'\"])/si');
1783          array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
1784                     "\\1$secremoveimg\\2");
1785          array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
1786                     '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
1787          array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
1788                     "url(\\1$secremoveimg\\2)");
1789     }
1790
1791     $add_attr_to_tag = Array(
1792                              "/^a$/i" => Array('target'=>'"_new"')
1793                              );
1794     $trusted = sq_sanitize($body,
1795                            $tag_list,
1796                            $rm_tags_with_content,
1797                            $self_closing_tags,
1798                            $force_tag_closing,
1799                            $rm_attnames,
1800                            $bad_attvals,
1801                            $add_attr_to_tag,
1802                            $message,
1803                            $id,
1804                            $mailbox
1805                            );
1806     if (preg_match("|$secremoveimg|i", $trusted)){
1807         $has_unsafe_images = true;
1808     }
1809     return $trusted;
1810 }
1811
1812 ?>