filter.test 32.2 KB
Newer Older
1 2 3
<?php
// $Id$

4
class FilterAdminTestCase extends DrupalWebTestCase {
5
  public static function getInfo() {
6
    return array(
7 8 9
      'name' => 'Filter administration functionality',
      'description' => 'Thoroughly test the administrative interface of the filter module.',
      'group' => 'Filter',
10 11 12 13 14 15
    );
  }

  /**
   * Test filter administration functionality.
   */
16
  function testFilterAdmin() {
17 18 19 20
    // URL filter.
    $first_filter = 2;
    // Line filter.
    $second_filter = 1;
21 22 23 24 25 26 27 28 29 30 31

    // Create users.
    $admin_user = $this->drupalCreateUser(array('administer filters'));
    $web_user = $this->drupalCreateUser(array('create page content'));
    $this->drupalLogin($admin_user);

    list($filtered, $full) = $this->checkFilterFormats();

    // Change default filter.
    $edit = array();
    $edit['default'] = $full;
32
    $this->drupalPost('admin/settings/formats', $edit, t('Save changes'));
33 34
    $this->assertText(t('Default format updated.'), t('Default filter updated successfully.'));

35
    $this->assertNoRaw('admin/settings/formats/delete/' . $full, t('Delete link not found.'));
36 37 38

    // Add an additional tag.
    $edit = array();
39
    $edit['allowed_html_1'] = '<a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd> <quote>';
40
    $this->drupalPost('admin/settings/formats/' . $filtered . '/configure', $edit, t('Save configuration'));
41 42 43 44
    $this->assertText(t('The configuration options have been saved.'), t('Allowed HTML tag added.'));

    $this->assertRaw(htmlentities($edit['allowed_html_1']), t('Tag displayed.'));

45
    $result = db_query('SELECT * FROM {cache_filter}')->fetchObject();
46 47 48 49
    $this->assertFalse($result, t('Cache cleared.'));

    // Reorder filters.
    $edit = array();
50 51
    $edit['weights[filter/' . $second_filter . ']'] = 1;
    $edit['weights[filter/' . $first_filter . ']'] = 2;
52
    $this->drupalPost('admin/settings/formats/' . $filtered . '/order', $edit, t('Save configuration'));
53 54
    $this->assertText(t('The filter ordering has been saved.'), t('Order saved successfully.'));

55
    $result = db_query('SELECT * FROM {filter} WHERE format = :format ORDER BY weight ASC', array(':format' => $filtered));
56
    $filters = array();
57
    foreach ($result as $filter) {
58 59 60 61 62 63 64 65 66 67
      if ($filter->delta == $second_filter || $filter->delta == $first_filter) {
        $filters[] = $filter;
      }
    }
    $this->assertTrue(($filters[0]->delta == $second_filter && $filters[1]->delta == $first_filter), t('Order confirmed.'));

    // Add filter.
    $edit = array();
    $edit['name'] = $this->randomName();
    $edit['roles[2]'] = TRUE;
68 69
    $edit['filters[filter/' . $second_filter . ']'] = TRUE;
    $edit['filters[filter/' . $first_filter . ']'] = TRUE;
70
    $this->drupalPost('admin/settings/formats/add', $edit, t('Save configuration'));
71
    $this->assertRaw(t('Added text format %format.', array('%format' => $edit['name'])), t('New filter created.'));
72 73 74 75 76 77

    $format = $this->getFilter($edit['name']);
    $this->assertNotNull($format, t('Format found in database.'));

    if ($format !== NULL) {
      $this->assertFieldByName('roles[2]', '', t('Role found.'));
78 79
      $this->assertFieldByName('filters[filter/' . $second_filter . ']', '', t('Line break filter found.'));
      $this->assertFieldByName('filters[filter/' . $first_filter . ']', '', t('Url filter found.'));
80 81

      // Delete new filter.
82
      $this->drupalPost('admin/settings/formats/delete/' . $format->format, array(), t('Delete'));
83
      $this->assertRaw(t('Deleted text format %format.', array('%format' => $edit['name'])), t('Format successfully deleted.'));
84 85 86 87 88
    }

    // Change default filter back.
    $edit = array();
    $edit['default'] = $filtered;
89
    $this->drupalPost('admin/settings/formats', $edit, t('Save changes'));
90 91
    $this->assertText(t('Default format updated.'), t('Default filter updated successfully.'));

92
    $this->assertNoRaw('admin/settings/formats/delete/' . $filtered, t('Delete link not found.'));
93 94 95 96

    // Allow authenticated users on full HTML.
    $edit = array();
    $edit['roles[2]'] = TRUE;
97
    $this->drupalPost('admin/settings/formats/' . $full, $edit, t('Save configuration'));
98
    $this->assertText(t('The text format settings have been updated.'), t('Full HTML format successfully updated.'));
99 100 101 102 103 104

    // Switch user.
    $this->drupalLogout();
    $this->drupalLogin($web_user);

    $this->drupalGet('node/add/page');
105
    $this->assertRaw('<option value="' . $full . '">Full HTML</option>', t('Full HTML filter accessible.'));
106

107
    // Use filtered HTML and see if it removes tags that are not allowed.
108 109 110 111 112
    $body = $this->randomName();
    $extra_text = 'text';

    $edit = array();
    $edit['title'] = $this->randomName();
113 114
    $edit['body[0][value]'] = $body . '<random>' . $extra_text . '</random>';
    $edit['body[0][value_format]'] = $filtered;
115 116 117
    $this->drupalPost('node/add/page', $edit, t('Save'));
    $this->assertRaw(t('Page %title has been created.', array('%title' => $edit['title'])), t('Filtered node created.'));

118
    $node = $this->drupalGetNodeByTitle($edit['title']);
119 120
    $this->assertTrue($node, t('Node found in database.'));

121
    $this->drupalGet('node/' . $node->nid);
122 123 124 125 126 127 128
    $this->assertText($body . $extra_text, t('Filter removed invalid tag.'));

    // Switch user.
    $this->drupalLogout();
    $this->drupalLogin($admin_user);

    // Clean up.
129
    // Allowed tags.
130 131
    $edit = array();
    $edit['allowed_html_1'] = '<a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd>';
132
    $this->drupalPost('admin/settings/formats/' . $filtered . '/configure', $edit, t('Save configuration'));
133 134
    $this->assertText(t('The configuration options have been saved.'), t('Changes reverted.'));

135
    // Full HTML.
136 137
    $edit = array();
    $edit['roles[2]'] = FALSE;
138
    $this->drupalPost('admin/settings/formats/' . $full, $edit, t('Save configuration'));
139
    $this->assertText(t('The text format settings have been updated.'), t('Full HTML format successfully reverted.'));
140

141
    // Filter order.
142
    $edit = array();
143 144
    $edit['weights[filter/' . $second_filter . ']'] = 2;
    $edit['weights[filter/' . $first_filter . ']'] = 1;
145
    $this->drupalPost('admin/settings/formats/' . $filtered . '/order', $edit, t('Save configuration'));
146 147 148 149 150 151
    $this->assertText(t('The filter ordering has been saved.'), t('Order successfully reverted.'));
  }

  /**
   * Query the database to get the two basic formats.
   *
152 153
   * @return
   *   An array containing filtered and full filter ids.
154 155
   */
  function checkFilterFormats() {
156
    $result = db_query('SELECT format, name FROM {filter_format}');
157 158 159

    $filtered = -1;
    $full = -1;
160
    foreach ($result as $format) {
161 162 163
      if ($format->name == 'Filtered HTML') {
        $filtered = $format->format;
      }
164
      elseif ($format->name == 'Full HTML') {
165 166 167 168 169 170 171 172 173 174
        $full = $format->format;
      }
    }

    return array($filtered, $full);
  }

  /**
   * Get filter by name.
   *
175 176 177 178
   * @param $name
   *   Name of filter to find.
   * @return
   *   A filter object.
179 180
   */
  function getFilter($name) {
181
    return db_query("SELECT * FROM {filter_format} WHERE name = :name", array(':name' => $name))->fetchObject();
182 183
  }
}
184 185 186 187

class FilterTestCase extends DrupalWebTestCase {
  protected $format;

188
  public static function getInfo() {
189
    return array(
190 191 192
      'name' => 'Core filters',
      'description' => 'Filter each filter individually: Convert URLs into links, Convert line breaks, Correct broken HTML, Escape all HTML, Limit allowed HTML tags.',
      'group' => 'Filter',
193 194 195 196 197
    );
  }

  function setUp() {
    parent::setUp();
198 199 200

    $admin_user = $this->drupalCreateUser(array('administer filters', 'create page content'));
    $this->drupalLogin($admin_user);
201 202 203
  }

  /**
204
   * Test the line break filter.
205 206
   */
  function testLineBreakFilter() {
207 208 209 210 211 212 213 214 215
    // Single line breaks should be changed to <br /> tags, while paragraphs
    // separated with double line breaks should be enclosed with <p></p> tags.
    $f = _filter_autop("aaa\nbbb\n\nccc");
    $this->assertEqual(str_replace("\n", '', $f), "<p>aaa<br />bbb</p><p>ccc</p>", t('Line breaking basic case.'));

    // Text within some contexts should not be processed.
    $f = _filter_autop("<script>aaa\nbbb\n\nccc</script>");
    $this->assertEqual($f, "<script>aaa\nbbb\n\nccc</script>", t('Line breaking -- do not break scripts.'));

216 217
    $f = _filter_autop('<p><div>  </div></p>');
    $this->assertEqual(substr_count($f, '<p>'), substr_count($f, '</p>'), t('Make sure line breaking produces matching paragraph tags.'));
Dries's avatar
Dries committed
218

219 220 221 222 223
    $f = _filter_autop('<div><p>  </p></div>');
    $this->assertEqual(substr_count($f, '<p>'), substr_count($f, '</p>'), t('Make sure line breaking produces matching paragraph tags.'));

    $f = _filter_autop('<blockquote><pre>aaa</pre></blockquote>');
    $this->assertEqual(substr_count($f, '<p>'), substr_count($f, '</p>'), t('Make sure line breaking produces matching paragraph tags.'));
224 225 226 227

    $limit = max(ini_get('pcre.backtrack_limit'), ini_get('pcre.recursion_limit'));
    $f = _filter_autop($this->randomName($limit));
    $this->assertNotEqual($f, '', t('Make sure line breaking can process long strings.'));
228 229 230
  }

  /**
231
   * Test limiting allowed tags, XSS prevention and adding 'nofollow' to links.
232 233 234
   *
   * XSS tests assume that script is dissallowed on default and src is allowed
   * on default, but on* and style are dissallowed.
235 236 237 238
   *
   * Script injection vectors mostly adopted from http://ha.ckers.org/xss.html.
   *
   * Relevant CVEs:
239 240
   * - CVE-2002-1806, ~CVE-2005-0682, ~CVE-2005-2106, CVE-2005-3973,
   *   CVE-2006-1226 (= rev. 1.112?), CVE-2008-0273, CVE-2008-3740.
241 242
   */
  function testHtmlFilter() {
243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274
    // Tag stripping, different ways to work around removal of HTML tags.
    $f = filter_xss('<script>alert(0)</script>');
    $this->assertNoNormalized($f, 'script', t('HTML tag stripping -- simple script without special characters.'));

    $f = filter_xss('<script src="http://www.example.com" />');
    $this->assertNoNormalized($f, 'script', t('HTML tag stripping -- empty script with source.'));

    $f = filter_xss('<ScRipt sRc=http://www.example.com/>');
    $this->assertNoNormalized($f, 'script', t('HTML tag stripping evasion -- varying case.'));

    $f = filter_xss("<script\nsrc\n=\nhttp://www.example.com/\n>");
    $this->assertNoNormalized($f, 'script', t('HTML tag stripping evasion -- multiline tag.'));

    $f = filter_xss('<script/a src=http://www.example.com/a.js></script>');
    $this->assertNoNormalized($f, 'script', t('HTML tag stripping evasion -- non whitespace character after tag name.'));

    $f = filter_xss('<script/src=http://www.example.com/a.js></script>');
    $this->assertNoNormalized($f, 'script', t('HTML tag stripping evasion -- no space between tag and attribute.'));

    // Null between < and tag name works at least with IE6.
    $f = filter_xss("<\0scr\0ipt>alert(0)</script>");
    $this->assertNoNormalized($f, 'ipt', t('HTML tag stripping evasion -- breaking HTML with nulls.'));

    $f = filter_xss("<scrscriptipt src=http://www.example.com/a.js>");
    $this->assertNoNormalized($f, 'script', t('HTML tag stripping evasion -- filter just removing "script".'));

    $f = filter_xss('<<script>alert(0);//<</script>');
    $this->assertNoNormalized($f, 'script', t('HTML tag stripping evasion -- double opening brackets.'));

    $f = filter_xss('<script src=http://www.example.com/a.js?<b>');
    $this->assertNoNormalized($f, 'script', t('HTML tag stripping evasion -- no closing tag.'));

275 276
    // DRUPAL-SA-2008-047: This doesn't seem exploitable, but the filter should
    // work consistently.
277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326
    $f = filter_xss('<script>>');
    $this->assertNoNormalized($f, 'script', t('HTML tag stripping evasion -- double closing tag.'));

    $f = filter_xss('<script src=//www.example.com/.a>');
    $this->assertNoNormalized($f, 'script', t('HTML tag stripping evasion -- no scheme or ending slash.'));

    $f = filter_xss('<script src=http://www.example.com/.a');
    $this->assertNoNormalized($f, 'script', t('HTML tag stripping evasion -- no closing bracket.'));

    $f = filter_xss('<script src=http://www.example.com/ <');
    $this->assertNoNormalized($f, 'script', t('HTML tag stripping evasion -- opening instead of closing bracket.'));

    $f = filter_xss('<nosuchtag attribute="newScriptInjectionVector">');
    $this->assertNoNormalized($f, 'nosuchtag', t('HTML tag stripping evasion -- unknown tag.'));

    $f = filter_xss('<?xml:namespace ns="urn:schemas-microsoft-com:time">');
    $this->assertTrue(stripos($f, '<?xml') === FALSE, t('HTML tag stripping evasion -- starting with a question sign (processing instructions).'));

    $f = filter_xss('<t:set attributeName="innerHTML" to="&lt;script defer&gt;alert(0)&lt;/script&gt;">');
    $this->assertNoNormalized($f, 't:set', t('HTML tag stripping evasion -- colon in the tag name (namespaces\' tricks).'));

    $f = filter_xss('<img """><script>alert(0)</script>', array('img'));
    $this->assertNoNormalized($f, 'script', t('HTML tag stripping evasion -- a malformed image tag.'));

    $f = filter_xss('<blockquote><script>alert(0)</script></blockquote>', array('blockquote'));
    $this->assertNoNormalized($f, 'script', t('HTML tag stripping evasion -- script in a blockqoute.'));

    $f = filter_xss("<!--[if true]><script>alert(0)</script><![endif]-->");
    $this->assertNoNormalized($f, 'script', t('HTML tag stripping evasion -- script within a comment.'));

    // Dangerous attributes removal.
    $f = filter_xss('<p onmouseover="http://www.example.com/">', array('p'));
    $this->assertNoNormalized($f, 'onmouseover', t('HTML filter attributes removal -- events, no evasion.'));

    $f = filter_xss('<li style="list-style-image: url(javascript:alert(0))">', array('li'));
    $this->assertNoNormalized($f, 'style', t('HTML filter attributes removal -- style, no evasion.'));

    $f = filter_xss('<img onerror   =alert(0)>', array('img'));
    $this->assertNoNormalized($f, 'onerror', t('HTML filter attributes removal evasion -- spaces before equals sign.'));

    $f = filter_xss('<img onabort!#$%&()*~+-_.,:;?@[/|\]^`=alert(0)>', array('img'));
    $this->assertNoNormalized($f, 'onabort', t('HTML filter attributes removal evasion -- non alphanumeric characters before equals sign.'));

    $f = filter_xss('<img oNmediAError=alert(0)>', array('img'));
    $this->assertNoNormalized($f, 'onmediaerror', t('HTML filter attributes removal evasion -- varying case.'));

    // Works at least with IE6.
    $f = filter_xss("<img o\0nfocus\0=alert(0)>", array('img'));
    $this->assertNoNormalized($f, 'focus', t('HTML filter attributes removal evasion -- breaking with nulls.'));

327
    // Only whitelisted scheme names allowed in attributes.
328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394
    $f = filter_xss('<img src="javascript:alert(0)">', array('img'));
    $this->assertNoNormalized($f, 'javascript', t('HTML scheme clearing -- no evasion.'));

    $f = filter_xss('<img src=javascript:alert(0)>', array('img'));
    $this->assertNoNormalized($f, 'javascript', t('HTML scheme clearing evasion -- no quotes.'));

    // A bit like CVE-2006-0070.
    $f = filter_xss('<img src="javascript:confirm(0)">', array('img'));
    $this->assertNoNormalized($f, 'javascript', t('HTML scheme clearing evasion -- no alert ;)'));

    $f = filter_xss('<img src=`javascript:alert(0)`>', array('img'));
    $this->assertNoNormalized($f, 'javascript', t('HTML scheme clearing evasion -- grave accents.'));

    $f = filter_xss('<img dynsrc="javascript:alert(0)">', array('img'));
    $this->assertNoNormalized($f, 'javascript', t('HTML scheme clearing -- rare attribute.'));

    $f = filter_xss('<table background="javascript:alert(0)">', array('table'));
    $this->assertNoNormalized($f, 'javascript', t('HTML scheme clearing -- another tag.'));

    $f = filter_xss('<base href="javascript:alert(0);//">', array('base'));
    $this->assertNoNormalized($f, 'javascript', t('HTML scheme clearing -- one more attribute and tag.'));

    $f = filter_xss('<img src="jaVaSCriPt:alert(0)">', array('img'));
    $this->assertNoNormalized($f, 'javascript', t('HTML scheme clearing evasion -- varying case.'));

    $f = filter_xss('<img src=&#106;&#97;&#118;&#97;&#115;&#99;&#114;&#105;&#112;&#116;&#58;&#97;&#108;&#101;&#114;&#116;&#40;&#48;&#41;>', array('img'));
    $this->assertNoNormalized($f, 'javascript', t('HTML scheme clearing evasion -- UTF-8 decimal encoding.'));

    $f = filter_xss('<img src=&#00000106&#0000097&#00000118&#0000097&#00000115&#0000099&#00000114&#00000105&#00000112&#00000116&#0000058&#0000097&#00000108&#00000101&#00000114&#00000116&#0000040&#0000048&#0000041>', array('img'));
    $this->assertNoNormalized($f, 'javascript', t('HTML scheme clearing evasion -- long UTF-8 encoding.'));

    $f = filter_xss('<img src=&#x6A&#x61&#x76&#x61&#x73&#x63&#x72&#x69&#x70&#x74&#x3A&#x61&#x6C&#x65&#x72&#x74&#x28&#x30&#x29>', array('img'));
    $this->assertNoNormalized($f, 'javascript', t('HTML scheme clearing evasion -- UTF-8 hex encoding.'));

    $f = filter_xss("<img src=\"jav\tascript:alert(0)\">", array('img'));
    $this->assertNoNormalized($f, 'script', t('HTML scheme clearing evasion -- an embedded tab.'));

    $f = filter_xss('<img src="jav&#x09;ascript:alert(0)">', array('img'));
    $this->assertNoNormalized($f, 'script', t('HTML scheme clearing evasion -- an encoded, embedded tab.'));

    $f = filter_xss('<img src="jav&#x000000A;ascript:alert(0)">', array('img'));
    $this->assertNoNormalized($f, 'script', t('HTML scheme clearing evasion -- an encoded, embedded newline.'));

    // With &#xD; this test would fail, but the entity gets turned into
    // &amp;#xD;, so it's OK.
    $f = filter_xss('<img src="jav&#x0D;ascript:alert(0)">', array('img'));
    $this->assertNoNormalized($f, 'script', t('HTML scheme clearing evasion -- an encoded, embedded carriage return.'));

    $f = filter_xss("<img src=\"\n\n\nj\na\nva\ns\ncript:alert(0)\">", array('img'));
    $this->assertNoNormalized($f, 'cript', t('HTML scheme clearing evasion -- broken into many lines.'));

    $f = filter_xss("<img src=\"jav\0a\0\0cript:alert(0)\">", array('img'));
    $this->assertNoNormalized($f, 'cript', t('HTML scheme clearing evasion -- embedded nulls.'));

    $f = filter_xss('<img src=" &#14;  javascript:alert(0)">', array('img'));
    $this->assertNoNormalized($f, 'javascript', t('HTML scheme clearing evasion -- spaces and metacharacters before scheme.'));

    $f = filter_xss('<img src="vbscript:msgbox(0)">', array('img'));
    $this->assertNoNormalized($f, 'vbscript', t('HTML scheme clearing evasion -- another scheme.'));

    $f = filter_xss('<img src="nosuchscheme:notice(0)">', array('img'));
    $this->assertNoNormalized($f, 'nosuchscheme', t('HTML scheme clearing evasion -- unknown scheme.'));

    // Netscape 4.x javascript entities.
    $f = filter_xss('<br size="&{alert(0)}">', array('br'));
    $this->assertNoNormalized($f, 'alert', t('Netscape 4.x javascript entities.'));

395 396 397
    // DRUPAL-SA-2008-006: Invalid UTF-8, these only work as reflected XSS with
    // Internet Explorer 6.
    $f = filter_xss("<p arg=\"\xe0\">\" style=\"background-image: url(javascript:alert(0));\"\xe0<p>", array('p'));
398 399 400 401
    $this->assertNoNormalized($f, 'style', t('HTML filter -- invalid UTF-8.'));

    $f = filter_xss("\xc0aaa");
    $this->assertEqual($f, '', t('HTML filter -- overlong UTF-8 sequences.'));
402 403 404 405 406 407 408 409 410

    $f = filter_xss("Who&#039;s Online");
    $this->assertNormalized($f, "who's online", t('HTML filter -- html entity number'));

    $f = filter_xss("Who&amp;#039;s Online");
    $this->assertNormalized($f, "who&#039;s online", t('HTML filter -- encoded html entity number'));

    $f = filter_xss("Who&amp;amp;#039; Online");
    $this->assertNormalized($f, "who&amp;#039; online", t('HTML filter -- double encoded html entity number'));
411 412 413 414 415
  }

  /**
   * Test filter settings, defaults, access restrictions and similar.
   *
416 417 418 419
   * @todo This is for functions like filter_filter and check_markup, whose
   *   functionality is not completely focused on filtering. Some ideas:
   *   restricting formats according to user permissions, proper cache
   *   handling, defaults -- allowed tags/attributes/protocols.
420
   *
421 422
   * @todo It is possible to add script, iframe etc. to allowed tags, but this
   *   makes HTML filter completely ineffective.
423
   *
424 425
   * @todo Class, id, name and xmlns should be added to disallowed attributes,
   *   or better a whitelist approach should be used for that too.
426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449
   */
  function testFilter() {
    // Check that access restriction really works.

    // HTML filter is not able to secure some tags, these should never be
    // allowed.
    $f = filter_filter('process', 0, 'no_such_format', '<script />');
    $this->assertNoNormalized($f, 'script', t('HTML filter should always remove script tags.'));

    $f = filter_filter('process', 0, 'no_such_format', '<iframe />');
    $this->assertNoNormalized($f, 'iframe', t('HTML filter should always remove iframe tags.'));

    $f = filter_filter('process', 0, 'no_such_format', '<object />');
    $this->assertNoNormalized($f, 'object', t('HTML filter should always remove object tags.'));

    $f = filter_filter('process', 0, 'no_such_format', '<style />');
    $this->assertNoNormalized($f, 'style', t('HTML filter should always remove style tags.'));

    // Some tags make CSRF attacks easier, let the user take the risk herself.
    $f = filter_filter('process', 0, 'no_such_format', '<img />');
    $this->assertNoNormalized($f, 'img', t('HTML filter should remove img tags on default.'));

    $f = filter_filter('process', 0, 'no_such_format', '<input />');
    $this->assertNoNormalized($f, 'img', t('HTML filter should remove input tags on default.'));
450

451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499
    // Filtering content of some attributes is infeasible, these shouldn't be
    // allowed too.
    $f = filter_filter('process', 0, 'no_such_format', '<p style="display: none;" />');
    $this->assertNoNormalized($f, 'style', t('HTML filter should remove style attribute on default.'));

    $f = filter_filter('process', 0, 'no_such_format', '<p onerror="alert(0);" />');
    $this->assertNoNormalized($f, 'onerror', t('HTML filter should remove on* attributes on default.'));
  }

  /**
   * Test the spam deterrent.
   */
  function testNoFollowFilter() {
    variable_set('filter_html_nofollow_f', TRUE);

    // Test if the rel="nofollow" attribute is added, even if we try to prevent
    // it.
    $f = _filter_html('<a href="http://www.example.com/">text</a>', 'f');
    $this->assertNormalized($f, 'rel="nofollow"', t('Spam deterrent -- no evasion.'));

    $f = _filter_html('<A href="http://www.example.com/">text</a>', 'f');
    $this->assertNormalized($f, 'rel="nofollow"', t('Spam deterrent evasion -- capital A.'));

    $f = _filter_html("<a/href=\"http://www.example.com/\">text</a>", 'f');
    $this->assertNormalized($f, 'rel="nofollow"', t('Spam deterrent evasion -- non whitespace character after tag name.'));

    $f = _filter_html("<\0a\0 href=\"http://www.example.com/\">text</a>", 'f');
    $this->assertNormalized($f, 'rel="nofollow"', t('Spam deterrent evasion -- some nulls.'));

    $f = _filter_html('<!--[if true]><a href="http://www.example.com/">text</a><![endif]-->', 'f');
    $this->assertNormalized($f, 'rel="nofollow"', t('Spam deterrent evasion -- link within a comment.'));
  }

  /**
   * Test the loose, admin HTML filter.
   */
  function testAdminHtmlFilter() {
    // DRUPAL-SA-2008-044
    $f = filter_xss_admin('<object />');
    $this->assertNoNormalized($f, 'object', t('Admin HTML filter -- should not allow object tag.'));

    $f = filter_xss_admin('<script />');
    $this->assertNoNormalized($f, 'script', t('Admin HTML filter -- should not allow script tag.'));

    $f = filter_xss_admin('<style /><iframe /><frame /><frameset /><meta /><link /><embed /><applet /><param /><layer />');
    $this->assertEqual($f, '', t('Admin HTML filter -- should never allow some tags.'));
  }

  /**
500 501 502
   * Test the HTML escaping filter.
   *
   * Here we test only whether check_plain() does what it should.
503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584
   */
  function testNoHtmlFilter() {
    // Test that characters that have special meaning in XML are changed into
    // entities.
    $f = check_plain('<>&"');
    $this->assertEqual($f, '&lt;&gt;&amp;&quot;', t('No HTML filter basic test.'));

    // A single quote can also be used for evil things in some contexts.
    $f = check_plain('\'');
    $this->assertEqual($f, '&#039;', t('No HTML filter -- single quote.'));

    // Test that the filter is not fooled by different evasion techniques.
    $f = check_plain("\xc2\"");
    $this->assertEqual($f, '', t('No HTML filter -- invalid UTF-8.'));
  }

  /**
   * Test the URL filter.
   */
  function testUrlFilter() {
    variable_set('filter_url_length_f', 496);

    // Converting URLs.
    $f = _filter_url('http://www.example.com/', 'f');
    $this->assertEqual($f, '<a href="http://www.example.com/" title="http://www.example.com/">http://www.example.com/</a>', t('Converting URLs.'));

    $f = _filter_url('http://www.example.com/?a=1&b=2', 'f');
    $this->assertEqual($f, '<a href="http://www.example.com/?a=1&amp;b=2" title="http://www.example.com/?a=1&amp;b=2">http://www.example.com/?a=1&amp;b=2</a>', t('Converting URLs -- ampersands.'));

    $f = _filter_url('ftp://user:pass@ftp.example.com/dir1/dir2', 'f');
    $this->assertEqual($f, '<a href="ftp://user:pass@ftp.example.com/dir1/dir2" title="ftp://user:pass@ftp.example.com/dir1/dir2">ftp://user:pass@ftp.example.com/dir1/dir2</a>', t('Converting URLs -- FTP scheme.'));

    // Converting domain names.
    $f = _filter_url('www.example.com', 'f');
    $this->assertEqual($f, '<a href="http://www.example.com" title="www.example.com">www.example.com</a>', t('Converting domain names.'));

    $f = _filter_url('<li>www.example.com</li>', 'f');
    $this->assertEqual($f, '<li><a href="http://www.example.com" title="www.example.com">www.example.com</a></li>', t('Converting domain names -- domain in a list.'));

    $f = _filter_url('(www.example.com/dir?a=1&b=2#a)', 'f');
    $this->assertEqual($f, '(<a href="http://www.example.com/dir?a=1&amp;b=2#a" title="www.example.com/dir?a=1&amp;b=2#a">www.example.com/dir?a=1&amp;b=2#a</a>)', t('Converting domain names --  domain in parentheses.'));

    // Converting e-mail addresses.
    $f = _filter_url('johndoe@example.com', 'f');
    $this->assertEqual($f, '<a href="mailto:johndoe@example.com">johndoe@example.com</a>', t('Converting e-mail addresses.'));

    $f = _filter_url('aaa@sub.tv', 'f');
    $this->assertEqual($f, '<a href="mailto:aaa@sub.tv">aaa@sub.tv</a>', t('Converting e-mail addresses -- a short e-mail from Tuvalu.'));

    // URL trimming.
    variable_set('filter_url_length_f', 28);

    $f = _filter_url('http://www.example.com/d/ff.ext?a=1&b=2#a1', 'f');
    $this->assertNormalized($f, 'http://www.example.com/d/ff....', t('URL trimming.'));

    // Not breaking existing links.
    $f = _filter_url('<a href="http://www.example.com">www.example.com</a>', 'f');
    $this->assertEqual($f, '<a href="http://www.example.com">www.example.com</a>', t('Converting URLs -- do not break existing links.'));

    $f = _filter_url('<a href="foo">http://www.example.com</a>', 'f');
    $this->assertEqual($f, '<a href="foo">http://www.example.com</a>', t('Converting URLs -- do not break existing, relative links.'));

    // Addresses within some tags such as code or script should not be converted.
    $f = _filter_url('<code>http://www.example.com</code>', 'f');
    $this->assertEqual($f, '<code>http://www.example.com</code>', t('Converting URLs -- skip code contents.'));

    $f = _filter_url('<code><em>http://www.example.com</em></code>', 'f');
    $this->assertEqual($f, '<code><em>http://www.example.com</em></code>', t('Converting URLs -- really skip code contents.'));

    $f = _filter_url('<script>http://www.example.com</script>', 'f');
    $this->assertEqual($f, '<script>http://www.example.com</script>', t('Converting URLs -- do not process scripts.'));

    // Addresses in attributes should not be converted.
    $f = _filter_url('<p xmlns="http://www.example.com" />', 'f');
    $this->assertEqual($f, '<p xmlns="http://www.example.com" />', t('Converting URLs -- do not convert addresses in attributes.'));

    $f = _filter_url('<a title="Go to www.example.com" href="http://www.example.com">text</a>', 'f');
    $this->assertEqual($f, '<a title="Go to www.example.com" href="http://www.example.com">text</a>', t('Converting URLs -- do not break existing links with custom title attribute.'));

    // Even though a dot at the end of a URL can indicate a fully qualified
    // domain name, such usage is rare compared to using a link at the end
    // of a sentence, so remove the dot from the link.
585
    // @todo It can also be used at the end of a filename or a query string.
586 587 588 589 590 591 592 593 594 595 596 597 598
    $f = _filter_url('www.example.com.', 'f');
    $this->assertEqual($f, '<a href="http://www.example.com" title="www.example.com">www.example.com</a>.', t('Converting URLs -- do not recognize a dot at the end of a domain name (FQDNs).'));

    $f = _filter_url('http://www.example.com.', 'f');
    $this->assertEqual($f, '<a href="http://www.example.com" title="http://www.example.com">http://www.example.com</a>.', t('Converting URLs -- do not recognize a dot at the end of an URL (FQDNs).'));

    $f = _filter_url('www.example.com/index.php?a=.', 'f');
    $this->assertEqual($f, '<a href="http://www.example.com/index.php?a=" title="www.example.com/index.php?a=">www.example.com/index.php?a=</a>.', t('Converting URLs -- do forget about a dot at the end of a query string.'));
  }

  /**
   * Test the HTML corrector.
   *
599
   * @todo This test could really use some validity checking function.
600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617
   */
  function testHtmlCorrector() {
    // Tag closing.
    $f = _filter_htmlcorrector('<p>text');
    $this->assertEqual($f, '<p>text</p>', t('HTML corrector -- tag closing at the end of input.'));

    $f = _filter_htmlcorrector('<p>text<p><p>text');
    $this->assertEqual($f, '<p>text</p><p></p><p>text</p>', t('HTML corrector -- tag closing.'));

    $f = _filter_htmlcorrector("<ul><li>e1<li>e2");
    $this->assertEqual($f, "<ul><li>e1</li><li>e2</li></ul>", t('HTML corrector -- unclosed list tags.'));

    $f = _filter_htmlcorrector('<div id="d">content');
    $this->assertEqual($f, '<div id="d">content</div>', t('HTML corrector -- unclosed tag with attribute.'));

    // XHTML slash for empty elements.
    $f = _filter_htmlcorrector('<hr><br>');
    $this->assertEqual($f, '<hr /><br />', t('HTML corrector -- XHTML closing slash.'));
618 619 620 621 622 623 624 625
  }

  function createFormat($filter) {
    $edit = array(
      'name' => $this->randomName(),
      'roles[2]' => TRUE,
      'filters[filter/' . $filter . ']' => TRUE,
    );
626
    $this->drupalPost('admin/settings/filter/add', $edit, t('Save configuration'));
627
    return db_query("SELECT * FROM {filter_format} WHERE name = :name", array(':name' => $edit['name']))->fetchObject();
628 629 630 631
  }

  function deleteFormat($format) {
    if ($format !== NULL) {
632
      $this->drupalPost('admin/settings/formats/delete/' . $format->format, array(), t('Delete'));
633 634
    }
  }
635 636

  /**
637
   * Asserts that a text transformed to lowercase with HTML entities decoded does contains a given string.
638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660
   *
   * Otherwise fails the test with a given message, similar to all the
   * SimpleTest assert* functions.
   *
   * Note that this does not remove nulls, new lines and other characters that
   * could be used to obscure a tag or an attribute name.
   *
   * @param $haystack
   *   Text to look in.
   * @param $needle
   *   Lowercase, plain text to look for.
   * @param $message
   *   Message to display if failed.
   * @param $group
   *   The group this message belongs to, defaults to 'Other'.
   * @return
   *   TRUE on pass, FALSE on fail.
   */
  function assertNormalized($haystack, $needle, $message = '', $group = 'Other') {
    return $this->assertTrue(strpos(strtolower(decode_entities($haystack)), $needle) !== FALSE, $message, $group);
  }

  /**
661
   * Asserts that text transformed to lowercase with HTML entities decoded does not contain a given string.
662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682
   *
   * Otherwise fails the test with a given message, similar to all the
   * SimpleTest assert* functions.
   *
   * Note that this does not remove nulls, new lines, and other character that
   * could be used to obscure a tag or an attribute name.
   *
   * @param $haystack
   *   Text to look in.
   * @param $needle
   *   Lowercase, plain text to look for.
   * @param $message
   *   Message to display if failed.
   * @param $group
   *   The group this message belongs to, defaults to 'Other'.
   * @return
   *   TRUE on pass, FALSE on fail.
   */
  function assertNoNormalized($haystack, $needle, $message = '', $group = 'Other') {
    return $this->assertTrue(strpos(strtolower(decode_entities($haystack)), $needle) === FALSE, $message, $group);
  }
683
}