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