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