fix for identifying literals in mime_match_parentheses. cyrus sometimes
[squirrelmail.git] / functions / mime.php
CommitLineData
59177427 1<?php
2ba13803 2
35586184 3/**
4 * mime.php
5 *
15e6162e 6 * Copyright (c) 1999-2002 The SquirrelMail Project Team
35586184 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 */
b74ba498 14
35586184 15require_once('../functions/imap.php');
16require_once('../functions/attachment_common.php');
8beafbbc 17
35586184 18/** Setting up the objects that have the structure for the message **/
19class msg_header {
20 /** msg_header contains generic variables for values that **/
21 /** could be in a header. **/
b74ba498 22
88cb1b4d 23 var $type0 = '', $type1 = '', $boundary = '', $charset = '',
24 $encoding = '', $size = 0, $to = array(), $from = '', $date = '',
25 $cc = array(), $bcc = array(), $reply_to = '', $subject = '',
26 $id = 0, $mailbox = '', $description = '', $filename = '',
c9d78ab4 27 $entity_id = 0, $message_id = 0, $name = '', $priority = 3, $type = '';
35586184 28}
b74ba498 29
451f74a2 30class message {
31 /** message is the object that contains messages. It is a recursive
32 object in that through the $entities variable, it can contain
33 more objects of type message. See documentation in mime.txt for
34 a better description of how this works.
35 **/
77b88425 36 var $header = '', $entities = array();
37
451f74a2 38 function addEntity ($msg) {
39 $this->entities[] = $msg;
40 }
41}
8beafbbc 42
451f74a2 43/* --------------------------------------------------------------------------------- */
44/* MIME DECODING */
45/* --------------------------------------------------------------------------------- */
b74ba498 46
451f74a2 47/* This function gets the structure of a message and stores it in the "message" class.
48 * It will return this object for use with all relevant header information and
49 * fully parsed into the standard "message" object format.
50 */
51function mime_structure ($imap_stream, $header) {
52
451f74a2 53 $ssid = sqimap_session_id();
54 $lsid = strlen( $ssid );
55 $id = $header->id;
56 fputs ($imap_stream, "$ssid FETCH $id BODYSTRUCTURE\r\n");
57 //
58 // This should use sqimap_read_data instead of reading it itself
59 //
77b88425 60 $read = fgets ($imap_stream, 9216);
451f74a2 61 $bodystructure = '';
62 while ( substr($read, 0, $lsid) <> $ssid &&
63 !feof( $imap_stream ) ) {
64 $bodystructure .= $read;
77b88425 65 $read = fgets ($imap_stream, 9216);
451f74a2 66 }
67 $read = $bodystructure;
77b88425 68
451f74a2 69 // isolate the body structure and remove beginning and end parenthesis
70 $read = trim(substr ($read, strpos(strtolower($read), 'bodystructure') + 13));
c9d78ab4 71
451f74a2 72 $read = trim(substr ($read, 0, -1));
73 $end = mime_match_parenthesis(0, $read);
74 while ($end == strlen($read)-1) {
75 $read = trim(substr ($read, 0, -1));
76 $read = trim(substr ($read, 1));
77 $end = mime_match_parenthesis(0, $read);
78 }
77b88425 79
451f74a2 80 $msg = mime_parse_structure ($read, 0);
81 $msg->header = $header;
77b88425 82
451f74a2 83 return( $msg );
84}
b74ba498 85
451f74a2 86/* this starts the parsing of a particular structure. It is called recursively,
87 * so it can be passed different structures. It returns an object of type
88 * $message.
89 * First, it checks to see if it is a multipart message. If it is, then it
90 * handles that as it sees is necessary. If it is just a regular entity,
91 * then it parses it and adds the necessary header information (by calling out
92 * to mime_get_elements()
93 */
94function mime_parse_structure ($structure, $ent_id) {
164800ad 95 global $mailbox;
c9d78ab4 96 $properties = array();
451f74a2 97 $msg = new message();
98 if ($structure{0} == '(') {
c9d78ab4 99 $old_ent_id = $ent_id;
451f74a2 100 $ent_id = mime_new_element_level($ent_id);
101 $start = $end = -1;
102 do {
103 $start = $end+1;
104 $end = mime_match_parenthesis ($start, $structure);
164800ad 105
106 /* check if we are dealing with a new entity-level */
107 $i = strrpos($ent_id,'.');
108 if ($i>0) {
109 $ent = substr($ent_id, $i+1);
110 } else {
111 $ent = '';
112 }
c9d78ab4 113 /* add "forgotten" parent entities (alternative and relative) */
164800ad 114 if ($ent == '0') {
115 /* new entity levels have information about the type (type1) and
116 * the properties. This information is situated at the end of the
117 * structure string like for example (example between the brackets)
118 * [ "RELATED" ("BOUNDARY" "myboundary" "TYPE" "plain/html") ]
119 */
120
121 /* get the involved properties for parsing to mime_get_properties */
122 $startprop = strrpos($structure,'(');
123 $properties_str = substr($structure,$startprop);
124 $endprop = mime_match_parenthesis ($startprop, $structure);
125 $propstr = substr($structure, $startprop + 1, ($endprop - $startprop)-1);
126 /* cut off the used properties */
127 if ($startprop) {
128 $structure_end = substr($structure, $endprop+2);
129 $structure = trim(substr($structure,0,$startprop));
130 }
164800ad 131 /* get type1 */
132 $pos = strrpos($structure,' ');
008989da 133 if ($structure{$pos+1} =='(') $pos++;
134
164800ad 135 $type1 = strtolower(substr($structure, $pos+2, (count($structure)-2)));
164800ad 136 /* cut off type1 */
137 if ($pos && $startprop) {
138 $structure = trim(substr($structure, 0, $pos));
139 }
140
141 /* process the found information */
142 $properties = mime_get_props($properties, $properties_str);
c9d78ab4 143 if (count($properties)>0) {
144 $msg->header->entity_id = $old_ent_id;
145 $msg->header->type0 = 'multipart';
146 $msg->header->type1 = $type1;
164800ad 147 for ($i=0; $i < count($properties); $i++) {
148 $msg->header->{$properties[$i]['name']} = $properties[$i]['value'];
149 }
c9d78ab4 150 }
164800ad 151 $structure = $structure . ' ' . $structure_end;
152 }
153 $element = substr($structure, $start+1, ($end - $start)-1);
154 $ent_id = mime_increment_id ($ent_id);
155 $newmsg = mime_parse_structure ($element, $ent_id);
156 /* set mailbox in case of message/rfc822 entities */
157 if (isset($newmsg->header->type0) && isset($newmsg->header->type1)) {
158 if ($newmsg->header->type0 == 'message' && $newmsg->header->type1 == 'rfc822') {
159 $newmsg->header->mailbox=$mailbox;
c9d78ab4 160 }
161 }
164800ad 162 $msg->addEntity ($newmsg);
451f74a2 163
451f74a2 164 } while ($structure{$end+1} == '(');
165 } else {
166 // parse the elements
c9d78ab4 167 $msg = mime_get_element ($structure, $msg, $ent_id);
451f74a2 168 }
169 return $msg;
170}
e4a256af 171
164800ad 172
451f74a2 173/* Increments the element ID. An element id can look like any of
174 * the following: 1, 1.2, 4.3.2.4.1, etc. This function increments
175 * the last number of the element id, changing 1.2 to 1.3.
176 */
177function mime_increment_id ($id) {
178
77b88425 179 if (strpos($id, '.')) {
180 $first = substr($id, 0, strrpos($id, '.'));
181 $last = substr($id, strrpos($id, '.')+1);
451f74a2 182 $last++;
77b88425 183 $new = $first . '.' .$last;
451f74a2 184 } else {
185 $new = $id + 1;
186 }
77b88425 187
451f74a2 188 return $new;
189}
190
191/*
192 * See comment for mime_increment_id().
193 * This adds another level on to the entity_id changing 1.3 to 1.3.0
194 * NOTE: 1.3.0 is not a valid element ID. It MUST be incremented
195 * before it can be used. I left it this way so as not to have
196 * to make a special case if it is the first entity_id. It
197 * always increments it, and that works fine.
198 */
199function mime_new_element_level ($id) {
200
77b88425 201 if (!$id) {
202 $id = 0;
203 } else {
204 $id = $id . '.0';
205 }
451f74a2 206
77b88425 207 return( $id );
451f74a2 208}
209
210function mime_get_element (&$structure, $msg, $ent_id) {
211
212 $elem_num = 1;
213 $msg->header = new msg_header();
214 $msg->header->entity_id = $ent_id;
215 $properties = array();
451f74a2 216 while (strlen($structure) > 0) {
217 $structure = trim($structure);
218 $char = $structure{0};
219
220 if (strtolower(substr($structure, 0, 3)) == 'nil') {
221 $text = '';
222 $structure = substr($structure, 3);
223 } else if ($char == '"') {
224 // loop through until we find the matching quote, and return that as a string
225 $pos = 1;
226 $text = '';
227 while ( ($char = $structure{$pos} ) <> '"' && $pos < strlen($structure)) {
228 $text .= $char;
229 $pos++;
230 }
231 $structure = substr($structure, strlen($text) + 2);
795d42e2 232 } else if ($char == '{') {
233 /**
234 * loop through until we find the matching quote,
235 * and return that as a string
236 */
237 $pos = 1;
238 $len = '';
239 while (($char = $structure{$pos}) != '}'
240 && $pos < strlen($structure)) {
241 $len .= $char;
242 $pos++;
243 }
244 $structure = substr($structure, strlen($len) + 4);
245 $text = substr($structure, 0, $len);
246 $structure = substr($structure, $len + 1);
451f74a2 247 } else if ($char == '(') {
248 // comment me
249 $end = mime_match_parenthesis (0, $structure);
250 $sub = substr($structure, 1, $end-1);
251 $properties = mime_get_props($properties, $sub);
252 $structure = substr($structure, strlen($sub) + 2);
253 } else {
254 // loop through until we find a space or an end parenthesis
255 $pos = 0;
256 $char = $structure{$pos};
257 $text = '';
258 while ($char != ' ' && $char != ')' && $pos < strlen($structure)) {
259 $text .= $char;
260 $pos++;
261 $char = $structure{$pos};
262 }
263 $structure = substr($structure, strlen($text));
264 }
265
266 // This is where all the text parts get put into the header
267 switch ($elem_num) {
268 case 1:
269 $msg->header->type0 = strtolower($text);
270 break;
271 case 2:
272 $msg->header->type1 = strtolower($text);
273 break;
274 case 4: // Id
275 // Invisimail enclose images with <>
276 $msg->header->id = str_replace( '<', '', str_replace( '>', '', $text ) );
277 break;
278 case 5:
279 $msg->header->description = $text;
280 break;
281 case 6:
282 $msg->header->encoding = strtolower($text);
283 break;
284 case 7:
285 $msg->header->size = $text;
286 break;
287 default:
288 if ($msg->header->type0 == 'text' && $elem_num == 8) {
289 // This is a plain text message, so lets get the number of lines
290 // that it contains.
291 $msg->header->num_lines = $text;
292
293 } else if ($msg->header->type0 == 'message' && $msg->header->type1 == 'rfc822' && $elem_num == 8) {
294 // This is an encapsulated message, so lets start all over again and
295 // parse this message adding it on to the existing one.
296 $structure = trim($structure);
297 if ( $structure{0} == '(' ) {
298 $e = mime_match_parenthesis (0, $structure);
299 $structure = substr($structure, 0, $e);
300 $structure = substr($structure, 1);
301 $m = mime_parse_structure($structure, $msg->header->entity_id);
302
303 // the following conditional is there to correct a bug that wasn't
304 // incrementing the entity IDs correctly because of the special case
305 // that message/rfc822 is. This fixes it fine.
306 if (substr($structure, 1, 1) != '(')
307 $m->header->entity_id = mime_increment_id(mime_new_element_level($ent_id));
308
309 // Now we'll go through and reformat the results.
310 if ($m->entities) {
311 for ($i=0; $i < count($m->entities); $i++) {
312 $msg->addEntity($m->entities[$i]);
313 }
314 } else {
315 $msg->addEntity($m);
316 }
317 $structure = "";
318 }
319 }
320 break;
321 }
322 $elem_num++;
323 $text = "";
324 }
325 // loop through the additional properties and put those in the various headers
c9d78ab4 326 for ($i=0; $i < count($properties); $i++) {
327 $msg->header->{$properties[$i]['name']} = $properties[$i]['value'];
328 }
451f74a2 329
330 return $msg;
331}
332
333/*
334 * I did most of the MIME stuff yesterday (June 20, 2000), but I couldn't
335 * figure out how to do this part, so I decided to go to bed. I woke up
336 * in the morning and had a flash of insight. I went to the white-board
337 * and scribbled it out, then spent a bit programming it, and this is the
338 * result. Nothing complicated, but I think my brain was fried yesterday.
339 * Funny how that happens some times.
340 *
341 * This gets properties in a nested parenthesisized list. For example,
342 * this would get passed something like: ("attachment" ("filename" "luke.tar.gz"))
343 * This returns an array called $props with all paired up properties.
344 * It ignores the "attachment" for now, maybe that should change later
345 * down the road. In this case, what is returned is:
346 * $props[0]["name"] = "filename";
347 * $props[0]["value"] = "luke.tar.gz";
348 */
349function mime_get_props ($props, $structure) {
350
351 while (strlen($structure) > 0) {
352 $structure = trim($structure);
353 $char = $structure{0};
451f74a2 354 if ($char == '"') {
355 $pos = 1;
356 $tmp = '';
357 while ( ( $char = $structure{$pos} ) != '"' &&
358 $pos < strlen($structure)) {
359 $tmp .= $char;
360 $pos++;
361 }
362 $structure = trim(substr($structure, strlen($tmp) + 2));
363 $char = $structure{0};
364
365 if ($char == '"') {
366 $pos = 1;
367 $value = '';
368 while ( ( $char = $structure{$pos} ) != '"' &&
369 $pos < strlen($structure) ) {
370 $value .= $char;
371 $pos++;
372 }
c9d78ab4 373 $structure = trim(substr($structure, strlen($value) + 2));
451f74a2 374 $k = count($props);
375 $props[$k]['name'] = strtolower($tmp);
376 $props[$k]['value'] = $value;
c9d78ab4 377 if ($structure != '') {
378 mime_get_props($props, $structure);
379 } else {
380 return $props;
381 }
451f74a2 382 } else if ($char == '(') {
383 $end = mime_match_parenthesis (0, $structure);
384 $sub = substr($structure, 1, $end-1);
c9d78ab4 385 if (! isset($props))
386 $props = array();
387 $props = mime_get_props($props, $sub);
388 $structure = substr($structure, strlen($sub) + 2);
389 return $props;
451f74a2 390 }
451f74a2 391 } else if ($char == '(') {
392 $end = mime_match_parenthesis (0, $structure);
393 $sub = substr($structure, 1, $end-1);
394 $props = mime_get_props($props, $sub);
395 $structure = substr($structure, strlen($sub) + 2);
396 return $props;
397 } else {
398 return $props;
399 }
400 }
401}
402
403/*
404 * Matches parenthesis. It will return the position of the matching
405 * parenthesis in $structure. For instance, if $structure was:
406 * ("text" "plain" ("val1name", "1") nil ... )
407 * x x
408 * then this would return 42 to match up those two.
409 */
410function mime_match_parenthesis ($pos, $structure) {
411
412 $j = strlen( $structure );
413
ec41cfc7 414 /*
415 * ignore all extra characters
416 * If inside of a quoted string or literal, skip it -- Boundary IDs and other
417 * things can have ) in them.
418 */
419
451f74a2 420 if ( $structure{$pos} != '(' ) {
421 return( $j );
422 }
423
424 while ( $pos < $j ) {
425 $pos++;
426 if ($structure{$pos} == ')') {
8beafbbc 427 return $pos;
ec41cfc7 428 } elseif ($structure{$pos} == '"') { /* check for quoted string */
b74ba498 429 $pos++;
451f74a2 430 while ( $structure{$pos} != '"' &&
431 $pos < $j ) {
432 if (substr($structure, $pos, 2) == '\\"') {
b74ba498 433 $pos++;
451f74a2 434 } elseif (substr($structure, $pos, 2) == '\\\\') {
b74ba498 435 $pos++;
451f74a2 436 }
b74ba498 437 $pos++;
5ffe5a7e 438 }
ec41cfc7 439 } elseif ($structure{$pos} == '{') { /* check for literal */
440 $str = substr($structure, $pos);
441 $pos++;
442 if (preg_match("/^\{(\d+)\}.*/",$str,$reg)) {
443 $pos = $pos + strlen($reg[1]) + $reg[1] + 2;
444 }
451f74a2 445 } elseif ( $structure{$pos} == '(' ) {
8beafbbc 446 $pos = mime_match_parenthesis ($pos, $structure);
451f74a2 447 }
448 }
449 echo _("Error decoding mime structure. Report this as a bug!") . '<br>';
450 return( $pos );
451}
452
93f92f03 453function mime_fetch_body($imap_stream, $id, $ent_id) {
09a4bde3 454
455 /*
456 * do a bit of error correction. If we couldn't find the entity id, just guess
457 * that it is the first one. That is usually the case anyway.
458 */
459 if (!$ent_id) {
451f74a2 460 $ent_id = 1;
09a4bde3 461 }
6ab1bd9e 462 $cmd = "FETCH $id BODY[$ent_id]";
77b88425 463
93f92f03 464 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message);
77b88425 465 do {
e976319c 466 $topline = trim(array_shift( $data ));
467 } while( $topline && $topline[0] == '*' && !preg_match( '/\* [0-9]+ FETCH.*/i', $topline )) ;
451f74a2 468 $wholemessage = implode('', $data);
469 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
77b88425 470
451f74a2 471 $ret = substr( $wholemessage, 0, $regs[1] );
472 /*
473 There is some information in the content info header that could be important
474 in order to parse html messages. Let's get them here.
475 */
476 if ( $ret{0} == '<' ) {
1c72b151 477 $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message);
e5ea9327 478 /* BASE within HTML documents is illegal (see w3 spec)
479* $base = '';
480* $k = 10;
481* foreach( $data as $d ) {
482* if ( substr( $d, 0, 13 ) == 'Content-Base:' ) {
483* $j = strlen( $d );
484* $i = 13;
485* $base = '';
486* while ( $i < $j &&
487* ( !isNoSep( $d{$i} ) || $d{$i} == '"' ) )
488* $i++;
489* while ( $i < $j ) {
490* if ( isNoSep( $d{$i} ) )
491* $base .= $d{$i};
492* $i++;
493* }
494* $k = 0;
495* } elseif ( $k == 1 && !isnosep( $d{0} ) ) {
496* $base .= substr( $d, 1 );
497* }
498* $k++;
499* }
500* if ( $base <> '' ) {
501* $ret = "<base href=\"$base\">" . $ret;
502* }
7e235a1a 503* */
451f74a2 504 }
505 } else if (ereg('"([^"]*)"', $topline, $regs)) {
506 $ret = $regs[1];
507 } else {
508 global $where, $what, $mailbox, $passed_id, $startMessage;
e5ea9327 509 $par = 'mailbox=' . urlencode($mailbox) . "&amp;passed_id=$passed_id";
451f74a2 510 if (isset($where) && isset($what)) {
e5ea9327 511 $par .= '&amp;where='. urlencode($where) . "&amp;what=" . urlencode($what);
a3daaaf3 512 } else {
e5ea9327 513 $par .= "&amp;startMessage=$startMessage&amp;show_more=0";
451f74a2 514 }
e5ea9327 515 $par .= '&amp;response=' . urlencode($response) .
516 '&amp;message=' . urlencode($message).
517 '&amp;topline=' . urlencode($topline);
a019eeb8 518
346817d4 519 echo '<tt><br>' .
520 '<table width="80%"><tr>' .
521 '<tr><td colspan=2>' .
451f74a2 522 _("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!") .
346817d4 523 " <A HREF=\"../src/retrievalerror.php?$par\"><br>" .
524 _("Submit message") . '</A><BR>&nbsp;' .
525 '</td></tr>' .
526 '<td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
527 '<td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
528 '<td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
529 '<td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
530 "</table><BR></tt></font><hr>";
531
1c72b151 532 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message);
451f74a2 533 array_shift($data);
534 $wholemessage = implode('', $data);
a019eeb8 535
346817d4 536 $ret = $wholemessage;
a3daaaf3 537 }
451f74a2 538 return( $ret );
539}
d4467150 540
451f74a2 541function mime_print_body_lines ($imap_stream, $id, $ent_id, $encoding) {
542 // do a bit of error correction. If we couldn't find the entity id, just guess
543 // that it is the first one. That is usually the case anyway.
544 if (!$ent_id) {
545 $ent_id = 1;
546 }
547 $sid = sqimap_session_id();
548 // Don't kill the connection if the browser is over a dialup
549 // and it would take over 30 seconds to download it.
b7206e1d 550
551 // donīt call set_time_limit in safe mode.
552 if (!ini_get("safe_mode")) {
553 set_time_limit(0);
554 }
346817d4 555
451f74a2 556 fputs ($imap_stream, "$sid FETCH $id BODY[$ent_id]\r\n");
557 $cnt = 0;
558 $continue = true;
559 $read = fgets ($imap_stream,4096);
560 // This could be bad -- if the section has sqimap_session_id() . ' OK'
561 // or similar, it will kill the download.
562 while (!ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
563 if (trim($read) == ')==') {
564 $read1 = $read;
b74ba498 565 $read = fgets ($imap_stream,4096);
451f74a2 566 if (ereg("^".$sid." (OK|BAD|NO)(.*)$", $read, $regs)) {
567 return;
568 } else {
569 echo decodeBody($read1, $encoding) .
570 decodeBody($read, $encoding);
571 }
572 } else if ($cnt) {
573 echo decodeBody($read, $encoding);
b74ba498 574 }
451f74a2 575 $read = fgets ($imap_stream,4096);
576 $cnt++;
577 }
578}
beb9e459 579
451f74a2 580/* -[ END MIME DECODING ]----------------------------------------------------------- */
d4467150 581
aceb0d5c 582
d4467150 583
451f74a2 584/* This is the first function called. It decides if this is a multipart
585 message or if it should be handled as a single entity
586 */
587function decodeMime ($imap_stream, &$header) {
588 global $username, $key, $imapServerAddress, $imapPort;
589 return mime_structure ($imap_stream, $header);
590}
ea48eb25 591
451f74a2 592// This is here for debugging purposese. It will print out a list
593// of all the entity IDs that are in the $message object.
f0c4dc12 594
451f74a2 595function listEntities ($message) {
596if ($message) {
597 if ($message->header->entity_id)
598 echo "<tt>" . $message->header->entity_id . ' : ' . $message->header->type0 . '/' . $message->header->type1 . '<br>';
599 for ($i = 0; $message->entities[$i]; $i++) {
600 $msg = listEntities($message->entities[$i], $ent_id);
601 if ($msg)
602 return $msg;
603 }
604}
605}
f0c4dc12 606
451f74a2 607
608/* returns a $message object for a particular entity id */
609function getEntity ($message, $ent_id) {
610 if ($message) {
c9d78ab4 611 if ($message->header->entity_id == $ent_id && strlen($ent_id) == strlen($message->header->entity_id))
612 {
8beafbbc 613 return $message;
451f74a2 614 } else {
cd928157 615 for ($i = 0; isset($message->entities[$i]); $i++) {
451f74a2 616 $msg = getEntity ($message->entities[$i], $ent_id);
617 if ($msg) {
618 return $msg;
c9d78ab4 619 }
b1dadc61 620 }
451f74a2 621 }
622 }
623}
8beafbbc 624
451f74a2 625/*
626 * figures out what entity to display and returns the $message object
627 * for that entity.
628 */
a0eed05c 629function findDisplayEntity ($msg, $textOnly = true, $entity = array() ) {
451f74a2 630 global $show_html_default;
631
93f92f03 632 $found = false;
c9d78ab4 633 if ($msg) {
93f92f03 634 $type = $msg->header->type0.'/'.$msg->header->type1;
635 if ( $type == 'multipart/alternative') {
a0eed05c 636 $msg = findAlternativeEntity($msg, $textOnly);
93f92f03 637 if (count($msg->entities) == 0) {
638 $entity[] = $msg->header->entity_id;
639 } else {
640 $found = true;
008989da 641 $entity =findDisplayEntity($msg,$textOnly, $entity);
93f92f03 642 }
643 } else if ( $type == 'multipart/related') {
008989da 644 $msgs = findRelatedEntity($msg);
645 for ($i = 0; $i < count($msgs); $i++) {
646 $msg = $msgs[$i];
647 if (count($msg->entities) == 0) {
648 $entity[] = $msg->header->entity_id;
649 } else {
650 $found = true;
651 $entity =findDisplayEntity($msg,$textOnly, $entity);
652 }
93f92f03 653 }
654 } else if ( count($entity) == 0 &&
c9d78ab4 655 $msg->header->type0 == 'text' &&
656 ( $msg->header->type1 == 'plain' ||
657 $msg->header->type1 == 'html' ) &&
658 isset($msg->header->entity_id) ) {
93f92f03 659 if (count($msg->entities) == 0) {
660 $entity[] = $msg->header->entity_id;
661 }
662 }
663 $i = 0;
664 while ( isset($msg->entities[$i]) && count($entity) == 0 && !$found ) {
008989da 665 $entity = findDisplayEntity($msg->entities[$i], $textOnly, $entity);
93f92f03 666 $i++;
667 }
668 }
669 if ( !isset($entity[0]) ) {
670 $entity[]="";
a3daaaf3 671 }
451f74a2 672 return( $entity );
673}
b74ba498 674
451f74a2 675/* Shows the HTML version */
676function findDisplayEntityHTML ($message) {
8405ee35 677
451f74a2 678 if ( $message->header->type0 == 'text' &&
679 $message->header->type1 == 'html' &&
680 isset($message->header->entity_id)) {
681 return $message->header->entity_id;
682 }
683 for ($i = 0; isset($message->entities[$i]); $i ++) {
f0c4dc12 684 if ( $message->header->type0 == 'message' &&
685 $message->header->type1 == 'rfc822' &&
686 isset($message->header->entity_id)) {
687 return 0;
688 }
93f92f03 689
451f74a2 690 $entity = findDisplayEntityHTML($message->entities[$i]);
691 if ($entity != 0) {
692 return $entity;
693 }
694 }
c9d78ab4 695
451f74a2 696 return 0;
697}
698
a0eed05c 699function findAlternativeEntity ($message, $textOnly) {
93f92f03 700 global $show_html_default;
701 /* if we are dealing with alternative parts then we choose the best
702 * viewable message supported by SM.
703 */
a0eed05c 704 if ($show_html_default && !$textOnly) {
93f92f03 705 $alt_order = array ('text/plain','text/html');
706 } else {
707 $alt_order = array ('text/plain');
708 }
709 $best_view = 0;
008989da 710 $ent_id = 0;
93f92f03 711 $k = 0;
712 for ($i = 0; $i < count($message->entities); $i ++) {
713 $type = $message->entities[$i]->header->type0.'/'.$message->entities[$i]->header->type1;
714 if ($type == 'multipart/related') {
715 $type = $message->entities[$i]->header->type;
716 }
717 for ($j = $k; $j < count($alt_order); $j++) {
718 if ($alt_order[$j] == $type && $j > $best_view) {
719 $best_view = $j;
008989da 720 $ent_id = $i;
93f92f03 721 $k = $j;
722 }
723 }
724 }
008989da 725 return $message->entities[$ent_id];
93f92f03 726}
727
728function findRelatedEntity ($message) {
008989da 729 $msgs = array();
93f92f03 730 for ($i = 0; $i < count($message->entities); $i ++) {
731 $type = $message->entities[$i]->header->type0.'/'.$message->entities[$i]->header->type1;
732 if ($message->header->type == $type) {
008989da 733 $msgs[] = $message->entities[$i];
93f92f03 734 }
735 }
008989da 736 return $msgs;
93f92f03 737}
738
da4c66e8 739/*
740 * translateText
741 * Extracted from strings.php 23/03/2002
742 */
743
744function translateText(&$body, $wrap_at, $charset) {
745 global $where, $what; /* from searching */
746 global $color; /* color theme */
747
748 require_once('../functions/url_parser.php');
749
750 $body_ary = explode("\n", $body);
751 $PriorQuotes = 0;
752 for ($i=0; $i < count($body_ary); $i++) {
753 $line = $body_ary[$i];
754 if (strlen($line) - 2 >= $wrap_at) {
755 sqWordWrap($line, $wrap_at);
756 }
757 $line = charset_decode($charset, $line);
758 $line = str_replace("\t", ' ', $line);
759
760 parseUrl ($line);
761
762 $Quotes = 0;
763 $pos = 0;
764 $j = strlen( $line );
765
766 while ( $pos < $j ) {
767 if ($line[$pos] == ' ') {
768 $pos ++;
769 } else if (strpos($line, '&gt;', $pos) === $pos) {
770 $pos += 4;
771 $Quotes ++;
772 } else {
773 break;
774 }
775 }
776
777 if ($Quotes > 1) {
778 if (! isset($color[14])) {
779 $color[14] = '#FF0000';
780 }
781 $line = '<FONT COLOR="' . $color[14] . '">' . $line . '</FONT>';
782 } elseif ($Quotes) {
783 if (! isset($color[13])) {
784 $color[13] = '#800000';
785 }
786 $line = '<FONT COLOR="' . $color[13] . '">' . $line . '</FONT>';
787 }
788
789 $body_ary[$i] = $line;
790 }
791 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
792}
793
c9d78ab4 794/* debugfunction for looping through entities and displaying correct entities */
795function listMyEntities ($message) {
796
797if ($message) {
798 if ($message->header->entity_id) {
799 echo "<tt>" . $message->header->entity_id . ' : ' . $message->header->type0 . '/' . $message->header->type1 . '<br>';
800 }
801 if (!($message->header->type0 == 'message' && $message->header->type1 == 'rfc822')) {
802 if (isset($message->header->boundary) ) {
803 $ent_id = $message->header->entity_id;
804 $var = $message->header->boundary;
805 if ($var !='')
806 echo "<b>$ent_id boundary = $var</b><br>";
807 }
808 if (isset($message->header->type) ) {
809 $var = $message->header->type;
810 if ($var !='')
811 echo "<b>$ent_id type = $var</b><br>";
812 }
813 for ($i = 0; $message->entities[$i]; $i++) {
814 $msg = listMyEntities($message->entities[$i]);
815 }
816
817 if ($msg ) return $msg;
818 }
819}
820
821}
822
823
824
451f74a2 825/* This returns a parsed string called $body. That string can then
826be displayed as the actual message in the HTML. It contains
827everything needed, including HTML Tags, Attachments at the
828bottom, etc.
829*/
93f92f03 830function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num) {
451f74a2 831 // this if statement checks for the entity to show as the
832 // primary message. To add more of them, just put them in the
833 // order that is their priority.
008989da 834 global $startMessage, $username, $key, $imapServerAddress, $imapPort,
d03c24f4 835 $show_html_default, $has_unsafe_images, $view_unsafe_images, $sort;
836
837 $has_unsafe_images = 0;
7e235a1a 838
451f74a2 839 $id = $message->header->id;
f0c4dc12 840
451f74a2 841 $urlmailbox = urlencode($message->header->mailbox);
93f92f03 842
451f74a2 843 $body_message = getEntity($message, $ent_num);
844 if (($body_message->header->type0 == 'text') ||
845 ($body_message->header->type0 == 'rfc822')) {
93f92f03 846 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
847
451f74a2 848 $body = decodeBody($body, $body_message->header->encoding);
849 $hookResults = do_hook("message_body", $body);
850 $body = $hookResults[1];
451f74a2 851 // If there are other types that shouldn't be formatted, add
852 // them here
853 if ($body_message->header->type1 == 'html') {
854 if ( $show_html_default <> 1 ) {
a3daaaf3 855 $body = strip_tags( $body );
856 translateText($body, $wrap_at, $body_message->header->charset);
857 } else {
93f92f03 858 $body = magicHTML( $body, $id, $message );
a3daaaf3 859 }
451f74a2 860 } else {
9eea179c 861 translateText($body, $wrap_at, $body_message->header->charset);
451f74a2 862 }
e5ea9327 863 $body .= "<CENTER><SMALL><A HREF=\"../src/download.php?absolute_dl=true&amp;passed_id=$id&amp;passed_ent_id=$ent_num&amp;mailbox=$urlmailbox&amp;showHeaders=1\">". _("Download this as a file") ."</A></SMALL></CENTER><BR>";
7e235a1a 864 if ($has_unsafe_images) {
865 if ($view_unsafe_images) {
3d1c7564 866 $body .= "<CENTER><SMALL><A HREF=\"read_body.php?passed_id=$id&amp;mailbox=$urlmailbox&amp;sort=$sort&amp;startMessage=$startMessage&amp;show_more=0\">". _("Hide Unsafe Images") ."</A></SMALL></CENTER><BR>\n";
d03c24f4 867 } else {
3d1c7564 868 $body .= "<CENTER><SMALL><A HREF=\"read_body.php?passed_id=$id&amp;mailbox=$urlmailbox&amp;sort=$sort&amp;startMessage=$startMessage&amp;show_more=0&amp;view_unsafe_images=1\">". _("View Unsafe Images") ."</A></SMALL></CENTER><BR>\n";
d03c24f4 869 }
7e235a1a 870 }
77b88425 871
451f74a2 872 /** Display the ATTACHMENTS: message if there's more than one part **/
164800ad 873 if (isset($message->entities[1])) {
e22815be 874 /* Header-type alternative means we choose the best one to display
875 so don't show the alternatives as attachment. Header-type related
876 means that the attachments are already part of the related message.
877 */
878 if ($message->header->type1 !='related' && $message->header->type1 !='alternative') {
879 $body .= formatAttachments ($message, $ent_num, $message->header->mailbox, $id);
880 }
451f74a2 881 }
451f74a2 882 } else {
883 $body = formatAttachments ($message, -1, $message->header->mailbox, $id);
884 }
885 return ($body);
886}
b74ba498 887
451f74a2 888/*
889 * A recursive function that returns a list of attachments with links
890 * to where to download these attachments
891 */
77b88425 892function formatAttachments($message, $ent_id, $mailbox, $id) {
893 global $where, $what;
894 global $startMessage, $color;
895 static $ShownHTML = 0;
451f74a2 896
77b88425 897 $body = '';
898 if ($ShownHTML == 0) {
451f74a2 899
77b88425 900 $ShownHTML = 1;
451f74a2 901 $body .= "<TABLE WIDTH=\"100%\" CELLSPACING=0 CELLPADDING=2 BORDER=0 BGCOLOR=\"$color[0]\"><TR>\n" .
77b88425 902 "<TH ALIGN=\"left\" BGCOLOR=\"$color[9]\"><B>\n" .
903 _("Attachments") . ':' .
904 "</B></TH></TR><TR><TD>\n" .
905 "<TABLE CELLSPACING=0 CELLPADDING=1 BORDER=0>\n" .
906 formatAttachments($message, $ent_id, $mailbox, $id) .
907 "</TABLE></TD></TR></TABLE>";
908
909 } else if ($message) {
f0c4dc12 910 $header = $message->header;
911 $type0 = strtolower($header->type0);
912 $type1 = strtolower($header->type1);
2d0476d6 913 $name = '';
914 if (isset($header->name)) {
915 $name = decodeHeader($header->name);
916 }
1fbbe1d4 917 if ($type0 =='message' && $type1 == 'rfc822') {
f0c4dc12 918
919 $filename = decodeHeader($message->header->filename);
920 if (trim($filename) == '') {
921 if (trim($name) == '') {
922 $display_filename = 'untitled-[' . $message->header->entity_id . ']' ;
923 } else {
924 $display_filename = $name;
925 $filename = $name;
926 }
927 } else {
928 $display_filename = $filename;
929 }
930
931 $urlMailbox = urlencode($mailbox);
932 $ent = urlencode($message->header->entity_id);
933
934 $DefaultLink =
935 "../src/download.php?startMessage=$startMessage&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;passed_ent_id=$ent";
936 if ($where && $what) {
937 $DefaultLink .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
938 }
939 $Links['download link']['text'] = _("download");
940 $Links['download link']['href'] =
941 "../src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;passed_ent_id=$ent";
942 $ImageURL = '';
943
944 /* this executes the attachment hook with a specific MIME-type.
945 * if that doens't have results, it tries if there's a rule
946 * for a more generic type. */
947 $HookResults = do_hook("attachment $type0/$type1", $Links,
948 $startMessage, $id, $urlMailbox, $ent, $DefaultLink, $display_filename, $where, $what);
949 if(count($HookResults[1]) <= 1) {
950 $HookResults = do_hook("attachment $type0/*", $Links,
951 $startMessage, $id, $urlMailbox, $ent, $DefaultLink,
952 $display_filename, $where, $what);
953 }
954
955 $Links = $HookResults[1];
956 $DefaultLink = $HookResults[6];
957
958 $body .= '<TR><TD>&nbsp;&nbsp;</TD><TD>' .
959 "<A HREF=\"$DefaultLink\">$display_filename</A>&nbsp;</TD>" .
960 '<TD><SMALL><b>' . show_readable_size($message->header->size) .
961 '</b>&nbsp;&nbsp;</small></TD>' .
962 "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
963 '<TD><SMALL>';
964 if ($message->header->description) {
965 $body .= '<b>' . htmlspecialchars(_($message->header->description)) . '</b>';
966 }
967 $body .= '</SMALL></TD><TD><SMALL>&nbsp;';
968
969
970 $SkipSpaces = 1;
971 foreach ($Links as $Val) {
972 if ($SkipSpaces) {
973 $SkipSpaces = 0;
974 } else {
975 $body .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
976 }
977 $body .= '<a href="' . $Val['href'] . '">' . $Val['text'] . '</a>';
978 }
979
980 unset($Links);
77b88425 981
f0c4dc12 982 $body .= "</SMALL></TD></TR>\n";
983
984 return( $body );
985
986 } elseif (!$message->entities) {
77b88425 987
988 $type0 = strtolower($message->header->type0);
989 $type1 = strtolower($message->header->type1);
990 $name = decodeHeader($message->header->name);
991
992 if ($message->header->entity_id != $ent_id) {
993 $filename = decodeHeader($message->header->filename);
994 if (trim($filename) == '') {
995 if (trim($name) == '') {
996 if ( trim( $message->header->id ) == '' )
997 $display_filename = 'untitled-[' . $message->header->entity_id . ']' ;
998 else
999 $display_filename = 'cid: ' . $message->header->id;
1000 // $display_filename = 'untitled-[' . $message->header->entity_id . ']' ;
1001 } else {
1002 $display_filename = $name;
1003 $filename = $name;
1004 }
1005 } else {
1006 $display_filename = $filename;
1007 }
451f74a2 1008
77b88425 1009 $urlMailbox = urlencode($mailbox);
1010 $ent = urlencode($message->header->entity_id);
451f74a2 1011
77b88425 1012 $DefaultLink =
e5ea9327 1013 "../src/download.php?startMessage=$startMessage&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;passed_ent_id=$ent";
77b88425 1014 if ($where && $what) {
ed268948 1015 $DefaultLink = '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
77b88425 1016 }
1017 $Links['download link']['text'] = _("download");
1018 $Links['download link']['href'] =
e5ea9327 1019 "../src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;passed_ent_id=$ent";
77b88425 1020 $ImageURL = '';
1021
1022 /* this executes the attachment hook with a specific MIME-type.
1023 * if that doens't have results, it tries if there's a rule
1024 * for a more generic type. */
1025 $HookResults = do_hook("attachment $type0/$type1", $Links,
1026 $startMessage, $id, $urlMailbox, $ent, $DefaultLink,
1027 $display_filename, $where, $what);
1028 if(count($HookResults[1]) <= 1) {
1029 $HookResults = do_hook("attachment $type0/*", $Links,
1030 $startMessage, $id, $urlMailbox, $ent, $DefaultLink,
1031 $display_filename, $where, $what);
1032 }
451f74a2 1033
77b88425 1034 $Links = $HookResults[1];
1035 $DefaultLink = $HookResults[6];
1036
1037 $body .= '<TR><TD>&nbsp;&nbsp;</TD><TD>' .
1038 "<A HREF=\"$DefaultLink\">$display_filename</A>&nbsp;</TD>" .
1039 '<TD><SMALL><b>' . show_readable_size($message->header->size) .
1040 '</b>&nbsp;&nbsp;</small></TD>' .
1041 "<TD><SMALL>[ $type0/$type1 ]&nbsp;</SMALL></TD>" .
1042 '<TD><SMALL>';
1043 if ($message->header->description) {
1044 $body .= '<b>' . htmlspecialchars(_($message->header->description)) . '</b>';
1045 }
1046 $body .= '</SMALL></TD><TD><SMALL>&nbsp;';
b74ba498 1047
b74ba498 1048
77b88425 1049 $SkipSpaces = 1;
1050 foreach ($Links as $Val) {
1051 if ($SkipSpaces) {
1052 $SkipSpaces = 0;
1053 } else {
1054 $body .= '&nbsp;&nbsp;|&nbsp;&nbsp;';
1055 }
1056 $body .= '<a href="' . $Val['href'] . '">' . $Val['text'] . '</a>';
1057 }
1058
1059 unset($Links);
1060
1061 $body .= "</SMALL></TD></TR>\n";
1062 }
1063 } else {
1064 for ($i = 0; $i < count($message->entities); $i++) {
1065 $body .= formatAttachments($message->entities[$i], $ent_id, $mailbox, $id);
1066 }
451f74a2 1067 }
77b88425 1068 }
1069 return( $body );
451f74a2 1070}
b74ba498 1071
b74ba498 1072
451f74a2 1073/** this function decodes the body depending on the encoding type. **/
1074function decodeBody($body, $encoding) {
1075 $body = str_replace("\r\n", "\n", $body);
1076 $encoding = strtolower($encoding);
b74ba498 1077
451f74a2 1078 global $show_html_default;
4809f489 1079
56c094f5 1080 if ($encoding == 'quoted-printable' ||
1081 $encoding == 'quoted_printable') {
451f74a2 1082 $body = quoted_printable_decode($body);
4809f489 1083
7831268e 1084
451f74a2 1085 while (ereg("=\n", $body))
1086 $body = ereg_replace ("=\n", "", $body);
358f007e 1087
451f74a2 1088 } else if ($encoding == 'base64') {
1089 $body = base64_decode($body);
1090 }
b74ba498 1091
451f74a2 1092 // All other encodings are returned raw.
1093 return $body;
1094}
1095
1096/*
1097 * This functions decode strings that is encoded according to
1098 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
79e07c7e 1099 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
451f74a2 1100 */
1101function decodeHeader ($string, $utfencode=true) {
79e07c7e 1102 if (is_array($string)) {
1103 $string = implode("\n", $string);
1104 }
1105 $i = 0;
1106 while (preg_match('/^(.{' . $i . '})(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=/Ui',
1107 $string, $res)) {
1108 $prefix = $res[1];
1109 // Ignore white-space between consecutive encoded-words
1110 if (strspn($res[2], " \t") != strlen($res[2])) {
1111 $prefix .= $res[2];
1112 }
18aa168e 1113
79e07c7e 1114 if (ucfirst($res[4]) == 'B') {
1115 $replace = base64_decode($res[5]);
1116 } else {
1117 $replace = str_replace('_', ' ', $res[5]);
1118 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
1119 $replace);
1120 /* Only encode into entities by default. Some places
1121 don't need the encoding, like the compose form. */
1122 if ($utfencode) {
1123 $replace = charset_decode($res[3], $replace);
1124 }
18aa168e 1125 }
79e07c7e 1126 $string = $prefix . $replace . substr($string, strlen($res[0]));
1127 $i = strlen($prefix) + strlen($replace);
18aa168e 1128 }
79e07c7e 1129 return( $string );
451f74a2 1130}
1131
1132/*
1133 * Encode a string according to RFC 1522 for use in headers if it
1134 * contains 8-bit characters or anything that looks like it should
1135 * be encoded.
1136 */
1137function encodeHeader ($string) {
1138 global $default_charset;
793cc001 1139
451f74a2 1140 // Encode only if the string contains 8-bit characters or =?
1141 $j = strlen( $string );
1142 $l = strstr($string, '=?'); // Must be encoded ?
1143 $ret = '';
1144 for( $i=0; $i < $j; ++$i) {
f7b3ba37 1145 switch( $string{$i} ) {
1146 case '=':
b74ba498 1147 $ret .= '=3D';
1148 break;
451f74a2 1149 case '?':
b74ba498 1150 $ret .= '=3F';
1151 break;
451f74a2 1152 case '_':
b74ba498 1153 $ret .= '=5F';
1154 break;
451f74a2 1155 case ' ':
b74ba498 1156 $ret .= '_';
1157 break;
451f74a2 1158 default:
b74ba498 1159 $k = ord( $string{$i} );
451f74a2 1160 if ( $k > 126 ) {
b74ba498 1161 $ret .= sprintf("=%02X", $k);
1162 $l = TRUE;
1163 } else
1164 $ret .= $string{$i};
f7b3ba37 1165 }
451f74a2 1166 }
793cc001 1167
451f74a2 1168 if ( $l ) {
f7b3ba37 1169 $string = "=?$default_charset?Q?$ret?=";
451f74a2 1170 }
346817d4 1171
451f74a2 1172 return( $string );
1173}
b74ba498 1174
691a2d25 1175/* This function trys to locate the entity_id of a specific mime element */
451f74a2 1176
691a2d25 1177function find_ent_id( $id, $message ) {
1178 $ret = '';
1179 for ($i=0; $ret == '' && $i < count($message->entities); $i++) {
164800ad 1180 if (( $message->entities[$i]->header->type1 == 'alternative') ||
1181 ( $message->entities[$i]->header->type1 == 'related') ||
1182 ( $message->entities[$i]->header->type1 == 'mixed')) {
1183 $ret = find_ent_id( $id, $message->entities[$i] );
451f74a2 1184 } else {
691a2d25 1185 if ( strcasecmp( $message->entities[$i]->header->id, $id ) == 0 )
1186 $ret = $message->entities[$i]->header->entity_id;
a3daaaf3 1187 }
691a2d25 1188
451f74a2 1189 }
691a2d25 1190 return( $ret );
451f74a2 1191}
a3daaaf3 1192
691a2d25 1193/**
1194 ** HTMLFILTER ROUTINES
1195 */
451f74a2 1196
691a2d25 1197/**
1198 * This function returns the final tag out of the tag name, an array
1199 * of attributes, and the type of the tag. This function is called by
1200 * sq_sanitize internally.
1201 *
1202 * @param $tagname the name of the tag.
1203 * @param $attary the array of attributes and their values
1204 * @param $tagtype The type of the tag (see in comments).
1205 * @return a string with the final tag representation.
1206 */
1207function sq_tagprint($tagname, $attary, $tagtype){
1208 $me = "sq_tagprint";
1209 if ($tagtype == 2){
1210 $fulltag = '</' . $tagname . '>';
1211 } else {
1212 $fulltag = '<' . $tagname;
1213 if (is_array($attary) && sizeof($attary)){
1214 $atts = Array();
1215 while (list($attname, $attvalue) = each($attary)){
1216 array_push($atts, "$attname=$attvalue");
1217 }
1218 $fulltag .= ' ' . join(" ", $atts);
1219 }
1220 if ($tagtype == 3){
1221 $fulltag .= " /";
1222 }
1223 $fulltag .= ">";
451f74a2 1224 }
691a2d25 1225 return $fulltag;
451f74a2 1226}
a3daaaf3 1227
691a2d25 1228/**
1229 * A small helper function to use with array_walk. Modifies a by-ref
1230 * value and makes it lowercase.
1231 *
1232 * @param $val a value passed by-ref.
1233 * @return void since it modifies a by-ref value.
1234 */
1235function sq_casenormalize(&$val){
1236 $val = strtolower($val);
1237}
451f74a2 1238
691a2d25 1239/**
1240 * This function skips any whitespace from the current position within
1241 * a string and to the next non-whitespace value.
1242 *
1243 * @param $body the string
1244 * @param $offset the offset within the string where we should start
1245 * looking for the next non-whitespace character.
1246 * @return the location within the $body where the next
1247 * non-whitespace char is located.
1248 */
1249function sq_skipspace($body, $offset){
1250 $me = "sq_skipspace";
1251 preg_match("/^(\s*)/s", substr($body, $offset), $matches);
1252 if (sizeof($matches{1})){
1253 $count = strlen($matches{1});
1254 $offset += $count;
451f74a2 1255 }
691a2d25 1256 return $offset;
451f74a2 1257}
a3daaaf3 1258
691a2d25 1259/**
1260 * This function looks for the next character within a string. It's
1261 * really just a glorified "strpos", except it catches if failures
1262 * nicely.
1263 *
1264 * @param $body The string to look for needle in.
1265 * @param $offset Start looking from this position.
1266 * @param $needle The character/string to look for.
1267 * @return location of the next occurance of the needle, or
1268 * strlen($body) if needle wasn't found.
1269 */
1270function sq_findnxstr($body, $offset, $needle){
1271 $me = "sq_findnxstr";
1272 $pos = strpos($body, $needle, $offset);
1273 if ($pos === FALSE){
1274 $pos = strlen($body);
451f74a2 1275 }
691a2d25 1276 return $pos;
451f74a2 1277}
a3daaaf3 1278
691a2d25 1279/**
1280 * This function takes a PCRE-style regexp and tries to match it
1281 * within the string.
1282 *
1283 * @param $body The string to look for needle in.
1284 * @param $offset Start looking from here.
1285 * @param $reg A PCRE-style regex to match.
1286 * @return Returns a false if no matches found, or an array
1287 * with the following members:
1288 * - integer with the location of the match within $body
1289 * - string with whatever content between offset and the match
1290 * - string with whatever it is we matched
1291 */
1292function sq_findnxreg($body, $offset, $reg){
1293 $me = "sq_findnxreg";
1294 $matches = Array();
1295 $retarr = Array();
1296 preg_match("%^(.*?)($reg)%s", substr($body, $offset), $matches);
1297 if (!$matches{0}){
1298 $retarr = false;
1299 } else {
1300 $retarr{0} = $offset + strlen($matches{1});
1301 $retarr{1} = $matches{1};
1302 $retarr{2} = $matches{2};
1303 }
1304 return $retarr;
1305}
a3daaaf3 1306
691a2d25 1307/**
1308 * This function looks for the next tag.
1309 *
1310 * @param $body String where to look for the next tag.
1311 * @param $offset Start looking from here.
1312 * @return false if no more tags exist in the body, or
1313 * an array with the following members:
1314 * - string with the name of the tag
1315 * - array with attributes and their values
1316 * - integer with tag type (1, 2, or 3)
1317 * - integer where the tag starts (starting "<")
1318 * - integer where the tag ends (ending ">")
1319 * first three members will be false, if the tag is invalid.
1320 */
1321function sq_getnxtag($body, $offset){
1322 $me = "sq_getnxtag";
1323 if ($offset > strlen($body)){
1324 return false;
1325 }
1326 $lt = sq_findnxstr($body, $offset, "<");
1327 if ($lt == strlen($body)){
1328 return false;
1329 }
1330 /**
1331 * We are here:
1332 * blah blah <tag attribute="value">
1333 * \---------^
1334 */
1335 $pos = sq_skipspace($body, $lt+1);
1336 if ($pos >= strlen($body)){
1337 return Array(false, false, false, $lt, strlen($body));
1338 }
1339 /**
1340 * There are 3 kinds of tags:
1341 * 1. Opening tag, e.g.:
1342 * <a href="blah">
1343 * 2. Closing tag, e.g.:
1344 * </a>
1345 * 3. XHTML-style content-less tag, e.g.:
1346 * <img src="blah"/>
1347 */
1348 $tagtype = false;
1349 switch (substr($body, $pos, 1)){
1350 case "/":
1351 $tagtype = 2;
1352 $pos++;
1353 break;
1354 case "!":
1355 /**
1356 * A comment or an SGML declaration.
1357 */
1358 if (substr($body, $pos+1, 2) == "--"){
bb8d0799 1359 $gt = strpos($body, "-->", $pos);
691a2d25 1360 if ($gt === false){
1361 $gt = strlen($body);
bb8d0799 1362 } else {
1363 $gt += 2;
1364 }
691a2d25 1365 return Array(false, false, false, $lt, $gt);
1366 } else {
1367 $gt = sq_findnxstr($body, $pos, ">");
1368 return Array(false, false, false, $lt, $gt);
1369 }
1370 break;
1371 default:
1372 /**
1373 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1374 * later.
1375 */
1376 $tagtype = 1;
1377 break;
1378 }
a3daaaf3 1379
691a2d25 1380 $tag_start = $pos;
1381 $tagname = '';
1382 /**
1383 * Look for next [\W-_], which will indicate the end of the tag name.
1384 */
1385 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1386 if ($regary == false){
1387 return Array(false, false, false, $lt, strlen($body));
1388 }
1389 list($pos, $tagname, $match) = $regary;
1390 $tagname = strtolower($tagname);
1391
1392 /**
1393 * $match can be either of these:
1394 * '>' indicating the end of the tag entirely.
1395 * '\s' indicating the end of the tag name.
1396 * '/' indicating that this is type-3 xhtml tag.
1397 *
1398 * Whatever else we find there indicates an invalid tag.
1399 */
1400 switch ($match){
1401 case "/":
1402 /**
1403 * This is an xhtml-style tag with a closing / at the
1404 * end, like so: <img src="blah"/>. Check if it's followed
1405 * by the closing bracket. If not, then this tag is invalid
1406 */
1407 if (substr($body, $pos, 2) == "/>"){
1408 $pos++;
1409 $tagtype = 3;
1410 } else {
1411 $gt = sq_findnxstr($body, $pos, ">");
1412 $retary = Array(false, false, false, $lt, $gt);
1413 return $retary;
1414 }
1415 case ">":
1416 return Array($tagname, false, $tagtype, $lt, $pos);
1417 break;
1418 default:
1419 /**
1420 * Check if it's whitespace
1421 */
1422 if (preg_match("/\s/", $match)){
1423 } else {
1424 /**
1425 * This is an invalid tag! Look for the next closing ">".
1426 */
1427 $gt = sq_findnxstr($body, $offset, ">");
1428 return Array(false, false, false, $lt, $gt);
1429 }
1430 }
1431
1432 /**
1433 * At this point we're here:
1434 * <tagname attribute='blah'>
1435 * \-------^
1436 *
1437 * At this point we loop in order to find all attributes.
1438 */
1439 $attname = '';
1440 $atttype = false;
1441 $attary = Array();
1442
1443 while ($pos <= strlen($body)){
1444 $pos = sq_skipspace($body, $pos);
1445 if ($pos == strlen($body)){
1446 /**
1447 * Non-closed tag.
1448 */
1449 return Array(false, false, false, $lt, $pos);
1450 }
1451 /**
1452 * See if we arrived at a ">" or "/>", which means that we reached
1453 * the end of the tag.
1454 */
1455 $matches = Array();
164800ad 1456 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
c828931c 1457 /**
1458 * Yep. So we did.
1459 */
1460 $pos += strlen($matches{1});
1461 if ($matches{2} == "/>"){
1462 $tagtype = 3;
1463 $pos++;
1464 }
1465 return Array($tagname, $attary, $tagtype, $lt, $pos);
1466 }
a3daaaf3 1467
cca46357 1468 /**
691a2d25 1469 * There are several types of attributes, with optional
1470 * [:space:] between members.
1471 * Type 1:
1472 * attrname[:space:]=[:space:]'CDATA'
1473 * Type 2:
1474 * attrname[:space:]=[:space:]"CDATA"
1475 * Type 3:
1476 * attr[:space:]=[:space:]CDATA
1477 * Type 4:
1478 * attrname
cca46357 1479 *
691a2d25 1480 * We leave types 1 and 2 the same, type 3 we check for
1481 * '"' and convert to "&quot" if needed, then wrap in
1482 * double quotes. Type 4 we convert into:
1483 * attrname="yes".
cca46357 1484 */
691a2d25 1485 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1486 if ($regary == false){
1487 /**
1488 * Looks like body ended before the end of tag.
1489 */
1490 return Array(false, false, false, $lt, strlen($body));
cca46357 1491 }
691a2d25 1492 list($pos, $attname, $match) = $regary;
1493 $attname = strtolower($attname);
1494 /**
1495 * We arrived at the end of attribute name. Several things possible
1496 * here:
1497 * '>' means the end of the tag and this is attribute type 4
1498 * '/' if followed by '>' means the same thing as above
1499 * '\s' means a lot of things -- look what it's followed by.
1500 * anything else means the attribute is invalid.
1501 */
1502 switch($match){
1503 case "/":
1504 /**
1505 * This is an xhtml-style tag with a closing / at the
1506 * end, like so: <img src="blah"/>. Check if it's followed
1507 * by the closing bracket. If not, then this tag is invalid
1508 */
1509 if (substr($body, $pos, 2) == "/>"){
1510 $pos++;
1511 $tagtype = 3;
1512 } else {
50ab7972 1513 $gt = sq_findnxstr($body, $pos, ">");
691a2d25 1514 $retary = Array(false, false, false, $lt, $gt);
1515 return $retary;
451f74a2 1516 }
691a2d25 1517 case ">":
1518 $attary{$attname} = '"yes"';
1519 return Array($tagname, $attary, $tagtype, $lt, $pos);
1520 break;
1521 default:
1522 /**
1523 * Skip whitespace and see what we arrive at.
1524 */
1525 $pos = sq_skipspace($body, $pos);
1526 $char = substr($body, $pos, 1);
1527 /**
1528 * Two things are valid here:
1529 * '=' means this is attribute type 1 2 or 3.
1530 * \w means this was attribute type 4.
1531 * anything else we ignore and re-loop. End of tag and
1532 * invalid stuff will be caught by our checks at the beginning
1533 * of the loop.
1534 */
1535 if ($char == "="){
1536 $pos++;
1537 $pos = sq_skipspace($body, $pos);
1538 /**
1539 * Here are 3 possibilities:
1540 * "'" attribute type 1
1541 * '"' attribute type 2
1542 * everything else is the content of tag type 3
1543 */
1544 $quot = substr($body, $pos, 1);
1545 if ($quot == "'"){
1546 $regary = sq_findnxreg($body, $pos+1, "\'");
1547 if ($regary == false){
1548 return Array(false, false, false, $lt, strlen($body));
1549 }
1550 list($pos, $attval, $match) = $regary;
1551 $pos++;
1552 $attary{$attname} = "'" . $attval . "'";
1553 } else if ($quot == '"'){
1554 $regary = sq_findnxreg($body, $pos+1, '\"');
1555 if ($regary == false){
1556 return Array(false, false, false, $lt, strlen($body));
1557 }
1558 list($pos, $attval, $match) = $regary;
1559 $pos++;
1560 $attary{$attname} = '"' . $attval . '"';
1561 } else {
1562 /**
1563 * These are hateful. Look for \s, or >.
1564 */
1565 $regary = sq_findnxreg($body, $pos, "[\s>]");
1566 if ($regary == false){
1567 return Array(false, false, false, $lt, strlen($body));
7e235a1a 1568 }
691a2d25 1569 list($pos, $attval, $match) = $regary;
1570 /**
1571 * If it's ">" it will be caught at the top.
1572 */
1573 $attval = preg_replace("/\"/s", "&quot;", $attval);
1574 $attary{$attname} = '"' . $attval . '"';
451f74a2 1575 }
691a2d25 1576 } else if (preg_match("|[\w/>]|", $char)) {
1577 /**
1578 * That was attribute type 4.
1579 */
1580 $attary{$attname} = '"yes"';
451f74a2 1581 } else {
691a2d25 1582 /**
1583 * An illegal character. Find next '>' and return.
1584 */
1585 $gt = sq_findnxstr($body, $pos, ">");
1586 return Array(false, false, false, $lt, $gt);
451f74a2 1587 }
691a2d25 1588 }
1589 }
1590 /**
1591 * The fact that we got here indicates that the tag end was never
1592 * found. Return invalid tag indication so it gets stripped.
1593 */
1594 return Array(false, false, false, $lt, strlen($body));
1595}
1596
1597/**
1598 * This function checks attribute values for entity-encoded values
1599 * and returns them translated into 8-bit strings so we can run
1600 * checks on them.
1601 *
1602 * @param $attvalue A string to run entity check against.
1603 * @return Translated value.
1604 */
1605function sq_deent($attvalue){
1606 $me="sq_deent";
1607 /**
1608 * See if we have to run the checks first. All entities must start
1609 * with "&".
1610 */
1611 if (strpos($attvalue, "&") === false){
1612 return $attvalue;
1613 }
1614 /**
1615 * Check named entities first.
1616 */
1617 $trans = get_html_translation_table(HTML_ENTITIES);
1618 /**
1619 * Leave &quot; in, as it can mess us up.
1620 */
1621 $trans = array_flip($trans);
1622 unset($trans{"&quot;"});
1623 while (list($ent, $val) = each($trans)){
1624 $attvalue = preg_replace("/$ent*(\W)/si", "$val\\1", $attvalue);
1625 }
1626 /**
1627 * Now translate numbered entities from 1 to 255 if needed.
1628 */
1629 if (strpos($attvalue, "#") !== false){
1630 $omit = Array(34, 39);
1631 for ($asc=1; $asc<256; $asc++){
1632 if (!in_array($asc, $omit)){
1633 $chr = chr($asc);
1634 $attvalue = preg_replace("/\&#0*$asc;*(\D)/si", "$chr\\1",
1635 $attvalue);
1636 $attvalue = preg_replace("/\&#x0*".dechex($asc).";*(\W)/si",
1637 "$chr\\1", $attvalue);
a3daaaf3 1638 }
691a2d25 1639 }
1640 }
1641 return $attvalue;
1642}
1643
1644/**
1645 * This function runs various checks against the attributes.
1646 *
1647 * @param $tagname String with the name of the tag.
1648 * @param $attary Array with all tag attributes.
1649 * @param $rm_attnames See description for sq_sanitize
1650 * @param $bad_attvals See description for sq_sanitize
1651 * @param $add_attr_to_tag See description for sq_sanitize
1652 * @param $message message object
1653 * @param $id message id
1654 * @return Array with modified attributes.
1655 */
1656function sq_fixatts($tagname,
1657 $attary,
1658 $rm_attnames,
1659 $bad_attvals,
1660 $add_attr_to_tag,
1661 $message,
1662 $id
1663 ){
1664 $me = "sq_fixatts";
1665 while (list($attname, $attvalue) = each($attary)){
1666 /**
1667 * See if this attribute should be removed.
1668 */
1669 foreach ($rm_attnames as $matchtag=>$matchattrs){
1670 if (preg_match($matchtag, $tagname)){
1671 foreach ($matchattrs as $matchattr){
1672 if (preg_match($matchattr, $attname)){
1673 unset($attary{$attname});
1674 continue;
1675 }
451f74a2 1676 }
451f74a2 1677 }
691a2d25 1678 }
1679 /**
1680 * Remove any entities.
1681 */
1682 $attvalue = sq_deent($attvalue);
1683
1684 /**
1685 * Now let's run checks on the attvalues.
1686 * I don't expect anyone to comprehend this. If you do,
1687 * get in touch with me so I can drive to where you live and
1688 * shake your hand personally. :)
1689 */
1690 foreach ($bad_attvals as $matchtag=>$matchattrs){
1691 if (preg_match($matchtag, $tagname)){
1692 foreach ($matchattrs as $matchattr=>$valary){
1693 if (preg_match($matchattr, $attname)){
1694 /**
1695 * There are two arrays in valary.
1696 * First is matches.
1697 * Second one is replacements
1698 */
1699 list($valmatch, $valrepl) = $valary;
1700 $newvalue =
1701 preg_replace($valmatch, $valrepl, $attvalue);
1702 if ($newvalue != $attvalue){
1703 $attary{$attname} = $newvalue;
1704 }
1705 }
1706 }
451f74a2 1707 }
a3daaaf3 1708 }
691a2d25 1709 /**
1710 * Turn cid: urls into http-friendly ones.
1711 */
1712 if (preg_match("/^[\'\"]\s*cid:/si", $attvalue)){
1713 $attary{$attname} = sq_cid2http($message, $id, $attvalue);
1714 }
a3daaaf3 1715 }
691a2d25 1716 /**
1717 * See if we need to append any attributes to this tag.
1718 */
1719 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1720 if (preg_match($matchtag, $tagname)){
1721 $attary = array_merge($attary, $addattary);
1722 }
1723 }
1724 return $attary;
451f74a2 1725}
a3daaaf3 1726
691a2d25 1727/**
1728 * This function edits the style definition to make them friendly and
1729 * usable in squirrelmail.
1730 *
1731 * @param $message the message object
1732 * @param $id the message id
1733 * @param $content a string with whatever is between <style> and </style>
1734 * @return a string with edited content.
1735 */
1736function sq_fixstyle($message, $id, $content){
1737 global $view_unsafe_images;
1738 $me = "sq_fixstyle";
1739 /**
1740 * First look for general BODY style declaration, which would be
1741 * like so:
1742 * body {background: blah-blah}
1743 * and change it to .bodyclass so we can just assign it to a <div>
1744 */
1745 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1746 $secremoveimg = "../images/" . _("sec_remove_eng.png");
1747 /**
1748 * Fix url('blah') declarations.
1749 */
1750 $content = preg_replace("|url\(([\'\"])\s*\S+script\s*:.*?([\'\"])\)|si",
1751 "url(\\1$secremoveimg\\2)", $content);
1752 /**
1753 * Fix url('https*://.*) declarations but only if $view_unsafe_images
1754 * is false.
1755 */
1756 if (!$view_unsafe_images){
1757 $content = preg_replace("|url\(([\'\"])\s*https*:.*?([\'\"])\)|si",
1758 "url(\\1$secremoveimg\\2)", $content);
1759 }
1760
1761 /**
1762 * Fix urls that refer to cid:
1763 */
1764 while (preg_match("|url\(([\'\"]\s*cid:.*?[\'\"])\)|si", $content,
1765 $matches)){
1766 $cidurl = $matches{1};
1767 $httpurl = sq_cid2http($message, $id, $cidurl);
1768 $content = preg_replace("|url\($cidurl\)|si",
1769 "url($httpurl)", $content);
1770 }
a3daaaf3 1771
691a2d25 1772 /**
bb8d0799 1773 * Fix stupid css declarations which lead to vulnerabilities
691a2d25 1774 * in IE.
1775 */
bb8d0799 1776 $match = Array('/expression/si',
1777 '/behaviou*r/si',
1778 '/binding/si');
1779 $replace = Array('idiocy', 'idiocy', 'idiocy');
1780 $content = preg_replace($match, $replace, $content);
691a2d25 1781 return $content;
1782}
a3daaaf3 1783
691a2d25 1784/**
1785 * This function converts cid: url's into the ones that can be viewed in
1786 * the browser.
1787 *
1788 * @param $message the message object
1789 * @param $id the message id
1790 * @param $cidurl the cid: url.
1791 * @return a string with a http-friendly url
1792 */
1793function sq_cid2http($message, $id, $cidurl){
1794 /**
1795 * Get rid of quotes.
1796 */
1797 $quotchar = substr($cidurl, 0, 1);
1798 $cidurl = str_replace($quotchar, "", $cidurl);
1799 $cidurl = substr(trim($cidurl), 4);
1800 $httpurl = $quotchar . "../src/download.php?absolute_dl=true&amp;" .
1801 "passed_id=$id&amp;mailbox=" . urlencode($message->header->mailbox) .
1802 "&amp;passed_ent_id=" . find_ent_id($cidurl, $message) . $quotchar;
1803 return $httpurl;
1804}
1805
1806/**
1807 * This function changes the <body> tag into a <div> tag since we
1808 * can't really have a body-within-body.
1809 *
1810 * @param $attary an array of attributes and values of <body>
1811 * @return a modified array of attributes to be set for <div>
1812 */
1813function sq_body2div($attary){
1814 $me = "sq_body2div";
1815 $divattary = Array("class"=>"'bodyclass'");
1816 $bgcolor="#ffffff";
1817 $text="#000000";
1818 $styledef="";
1819 if (is_array($attary) && sizeof($attary) > 0){
1820 foreach ($attary as $attname=>$attvalue){
1821 $quotchar = substr($attvalue, 0, 1);
1822 $attvalue = str_replace($quotchar, "", $attvalue);
1823 switch ($attname){
1824 case "background":
1825 $styledef .= "background-image: url('$attvalue'); ";
1826 break;
1827 case "bgcolor":
1828 $styledef .= "background-color: $attvalue; ";
1829 break;
1830 case "text":
1831 $styledef .= "color: $attvalue; ";
1832 }
a3daaaf3 1833 }
691a2d25 1834 if (strlen($styledef) > 0){
1835 $divattary{"style"} = "\"$styledef\"";
1836 }
1837 }
1838 return $divattary;
1839}
a3daaaf3 1840
691a2d25 1841/**
1842 * This is the main function and the one you should actually be calling.
1843 * There are several variables you should be aware of an which need
1844 * special description.
1845 *
1846 * Since the description is quite lengthy, see it here:
1847 * http://www.mricon.com/html/phpfilter.html
1848 *
1849 * @param $body the string with HTML you wish to filter
1850 * @param $tag_list see description above
1851 * @param $rm_tags_with_content see description above
1852 * @param $self_closing_tags see description above
1853 * @param $force_tag_closing see description above
1854 * @param $rm_attnames see description above
1855 * @param $bad_attvals see description above
1856 * @param $add_attr_to_tag see description above
1857 * @param $message message object
1858 * @param $id message id
1859 * @return sanitized html safe to show on your pages.
1860 */
1861function sq_sanitize($body,
1862 $tag_list,
1863 $rm_tags_with_content,
1864 $self_closing_tags,
1865 $force_tag_closing,
1866 $rm_attnames,
1867 $bad_attvals,
1868 $add_attr_to_tag,
1869 $message,
1870 $id
1871 ){
1872 $me = "sq_sanitize";
1873 /**
1874 * Normalize rm_tags and rm_tags_with_content.
1875 */
1876 @array_walk($rm_tags, 'sq_casenormalize');
1877 @array_walk($rm_tags_with_content, 'sq_casenormalize');
1878 @array_walk($self_closing_tags, 'sq_casenormalize');
1879 /**
1880 * See if tag_list is of tags to remove or tags to allow.
1881 * false means remove these tags
1882 * true means allow these tags
1883 */
1884 $rm_tags = array_shift($tag_list);
1885 $curpos = 0;
1886 $open_tags = Array();
1887 $trusted = "<!-- begin sanitized html -->\n";
1888 $skip_content = false;
bb8d0799 1889 /**
1890 * Take care of netscape's stupid javascript entities like
1891 * &{alert('boo')};
1892 */
1893 $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
691a2d25 1894
1895 while (($curtag=sq_getnxtag($body, $curpos)) != FALSE){
1896 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
1897 $free_content = substr($body, $curpos, $lt-$curpos);
1898 /**
1899 * Take care of <style>
1900 */
1901 if ($tagname == "style" && $tagtype == 2){
1902 /**
1903 * This is a closing </style>. Edit the
1904 * content before we apply it.
1905 */
1906 $free_content = sq_fixstyle($message, $id, $free_content);
691a2d25 1907 }
1908 if ($skip_content == false){
1909 $trusted .= $free_content;
1910 } else {
1911 }
1912 if ($tagname != FALSE){
1913 if ($tagtype == 2){
1914 if ($skip_content == $tagname){
1915 /**
1916 * Got to the end of tag we needed to remove.
1917 */
1918 $tagname = false;
1919 $skip_content = false;
1920 } else {
1921 if ($skip_content == false){
c828931c 1922 if ($tagname == "body"){
1923 $tagname = "div";
691a2d25 1924 } else {
c828931c 1925 if (isset($open_tags{$tagname}) &&
1926 $open_tags{$tagname} > 0){
1927 $open_tags{$tagname}--;
1928 } else {
1929 $tagname = false;
1930 }
691a2d25 1931 }
1932 } else {
1933 }
1934 }
1935 } else {
1936 /**
1937 * $rm_tags_with_content
1938 */
1939 if ($skip_content == false){
1940 /**
1941 * See if this is a self-closing type and change
1942 * tagtype appropriately.
1943 */
1944 if ($tagtype == 1
1945 && in_array($tagname, $self_closing_tags)){
1946 $tagtype=3;
1947 }
1948 /**
1949 * See if we should skip this tag and any content
1950 * inside it.
1951 */
1952 if ($tagtype == 1 &&
1953 in_array($tagname, $rm_tags_with_content)){
1954 $skip_content = $tagname;
1955 } else {
1956 if (($rm_tags == false
1957 && in_array($tagname, $tag_list)) ||
1958 ($rm_tags == true &&
1959 !in_array($tagname, $tag_list))){
1960 $tagname = false;
1961 } else {
1962 if ($tagtype == 1){
1963 if (isset($open_tags{$tagname})){
1964 $open_tags{$tagname}++;
1965 } else {
1966 $open_tags{$tagname}=1;
1967 }
1968 }
1969 /**
1970 * This is where we run other checks.
1971 */
1972 if (is_array($attary) && sizeof($attary) > 0){
1973 $attary = sq_fixatts($tagname,
1974 $attary,
1975 $rm_attnames,
1976 $bad_attvals,
1977 $add_attr_to_tag,
1978 $message,
1979 $id
1980 );
1981 }
c828931c 1982 /**
1983 * Convert body into div.
1984 */
1985 if ($tagname == "body"){
1986 $tagname = "div";
1987 $attary = sq_body2div($attary, $message, $id);
1988 }
691a2d25 1989 }
1990 }
1991 } else {
1992 }
1993 }
1994 if ($tagname != false && $skip_content == false){
1995 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
1996 }
1997 } else {
1998 }
1999 $curpos = $gt+1;
a3daaaf3 2000 }
691a2d25 2001 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2002 if ($force_tag_closing == true){
2003 foreach ($open_tags as $tagname=>$opentimes){
2004 while ($opentimes > 0){
2005 $trusted .= '</' . $tagname . '>';
2006 $opentimes--;
2007 }
2008 }
2009 $trusted .= "\n";
2010 }
2011 $trusted .= "<!-- end sanitized html -->\n";
2012 return $trusted;
2013}
451f74a2 2014
691a2d25 2015/**
2016 * This is a wrapper function to call html sanitizing routines.
2017 *
2018 * @param $body the body of the message
2019 * @param $id the id of the message
2020 * @return a string with html safe to display in the browser.
2021 */
0b331da5 2022function magicHTML($body, $id, $message){
691a2d25 2023 global $attachment_common_show_images, $view_unsafe_images,
794b2d0a 2024 $has_unsafe_images;
691a2d25 2025 /**
2026 * Don't display attached images in HTML mode.
2027 */
2028 $attachment_common_show_images = false;
2029 $tag_list = Array(
2030 false,
2031 "object",
2032 "meta",
2033 "html",
2034 "head",
2035 "base"
2036 );
2037
2038 $rm_tags_with_content = Array(
2039 "script",
2040 "applet",
2041 "embed",
2042 "title"
2043 );
2044
2045 $self_closing_tags = Array(
2046 "img",
2047 "br",
2048 "hr",
2049 "input"
2050 );
2051
2052 $force_tag_closing = false;
2053
2054 $rm_attnames = Array(
2055 "/.*/" =>
2056 Array(
2057 "/target/si",
bb8d0799 2058 "/^on.*/si",
2059 "/^dynsrc/si",
2060 "/^data.*/si"
691a2d25 2061 )
2062 );
2063
2064 $secremoveimg = "../images/" . _("sec_remove_eng.png");
2065 $bad_attvals = Array(
2066 "/.*/" =>
2067 Array(
2068 "/^src|background|href|action/i" =>
2069 Array(
2070 Array(
2071 "|^([\'\"])\s*\.\./.*([\'\"])|si",
bb8d0799 2072 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
b6f2416e 2073 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2074 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
691a2d25 2075 ),
2076 Array(
2077 "\\1$secremoveimg\\2",
bb8d0799 2078 "\\1$secremoveimg\\2",
2079 "\\1$secremoveimg\\2",
2080 "\\1$secremoveimg\\2"
691a2d25 2081 )
2082 ),
2083 "/^style/si" =>
2084 Array(
2085 Array(
bb8d0799 2086 "/expression/si",
2087 "/binding/si",
2088 "/behaviou*r/si",
691a2d25 2089 "|url\(([\'\"])\s*\.\./.*([\'\"])\)|si",
2090 "/url\(([\'\"])\s*\S+script:.*([\'\"])\)/si"
2091 ),
2092 Array(
bb8d0799 2093 "idiocy",
2094 "idiocy",
2095 "idiocy",
691a2d25 2096 "url(\\1$secremoveimg\\2)",
2097 "url(\\1$secremoveimg\\2)"
2098 )
2099 )
2100 )
2101 );
2102 if (!$view_unsafe_images){
2103 /**
2104 * Remove any references to http/https if view_unsafe_images set
2105 * to false.
2106 */
43d678b6 2107 array_push($bad_attvals{'/.*/'}{'/^src|background|href|action/i'}[0],
2108 '/^([\'\"])\s*https*:.*([\'\"])/si');
2109 array_push($bad_attvals{'/.*/'}{'/^src|background|href|action/i'}[1],
2110 "\\1$secremoveimg\\2");
2111 array_push($bad_attvals{'/.*/'}{'/^style/si'}[0],
2112 '/url\(([\'\"])\s*https*:.*([\'\"])\)/si');
2113 array_push($bad_attvals{'/.*/'}{'/^style/si'}[1],
2114 "url(\\1$secremoveimg\\2)");
691a2d25 2115 }
451f74a2 2116
691a2d25 2117 $add_attr_to_tag = Array(
2118 "/^a$/si" => Array('target'=>'"_new"')
2119 );
2120 $trusted = sq_sanitize($body,
2121 $tag_list,
2122 $rm_tags_with_content,
2123 $self_closing_tags,
2124 $force_tag_closing,
2125 $rm_attnames,
2126 $bad_attvals,
2127 $add_attr_to_tag,
2128 $message,
2129 $id
2130 );
2131 if (preg_match("|$secremoveimg|si", $trusted)){
2132 $has_unsafe_images = true;
2133 }
2134 return $trusted;
451f74a2 2135}
691a2d25 2136?>