functions/mime.php

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