tests/phpunit/includes/SanitizerTest.php

   1 <?php
   2
   3 class SanitizerTest extends MediaWikiTestCase {
   4
   5         protected function setUp() {
   6                 parent::setUp();
   7
   8                 AutoLoader::loadClass( 'Sanitizer' );
   9         }
  10
  11         function testDecodeNamedEntities() {
  12                 $this->assertEquals(
  13                         "\xc3\xa9cole",
  14                         Sanitizer::decodeCharReferences( '&eacute;cole' ),
  15                         'decode named entities'
  16                 );
  17         }
  18
  19         function testDecodeNumericEntities() {
  20                 $this->assertEquals(
  21                         "\xc4\x88io bonas dans l'\xc3\xa9cole!",
  22                         Sanitizer::decodeCharReferences( "&#x108;io bonas dans l'&#233;cole!" ),
  23                         'decode numeric entities'
  24                 );
  25         }
  26
  27         function testDecodeMixedEntities() {
  28                 $this->assertEquals(
  29                         "\xc4\x88io bonas dans l'\xc3\xa9cole!",
  30                         Sanitizer::decodeCharReferences( "&#x108;io bonas dans l'&eacute;cole!" ),
  31                         'decode mixed numeric/named entities'
  32                 );
  33         }
  34
  35         function testDecodeMixedComplexEntities() {
  36                 $this->assertEquals(
  37                         "\xc4\x88io bonas dans l'\xc3\xa9cole! (mais pas &#x108;io dans l'&eacute;cole)",
  38                         Sanitizer::decodeCharReferences(
  39                                 "&#x108;io bonas dans l'&eacute;cole! (mais pas &amp;#x108;io dans l'&#38;eacute;cole)"
  40                         ),
  41                         'decode mixed complex entities'
  42                 );
  43         }
  44
  45         function testInvalidAmpersand() {
  46                 $this->assertEquals(
  47                         'a & b',
  48                         Sanitizer::decodeCharReferences( 'a & b' ),
  49                         'Invalid ampersand'
  50                 );
  51         }
  52
  53         function testInvalidEntities() {
  54                 $this->assertEquals(
  55                         '&foo;',
  56                         Sanitizer::decodeCharReferences( '&foo;' ),
  57                         'Invalid named entity'
  58                 );
  59         }
  60
  61         function testInvalidNumberedEntities() {
  62                 $this->assertEquals( UTF8_REPLACEMENT, Sanitizer::decodeCharReferences( "&#88888888888888;" ), 'Invalid numbered entity' );
  63         }
  64
  65         /**
  66          * @cover Sanitizer::removeHTMLtags
  67          * @dataProvider provideHtml5Tags
  68          *
  69          * @param String $tag Name of an HTML5 element (ie: 'video')
  70          * @param Boolean $escaped Wheter sanitizer let the tag in or escape it (ie: '&lt;video&gt;')
  71          */
  72         function testRemovehtmltagsOnHtml5Tags( $tag, $escaped ) {
  73                 $this->setMwGlobals( array(
  74                         # Enable HTML5 mode
  75                         'wgHtml5' => true,
  76                         'wgUseTidy' => false
  77                 ));
  78
  79                 if( $escaped ) {
  80                         $this->assertEquals( "&lt;$tag&gt;",
  81                                 Sanitizer::removeHTMLtags( "<$tag>" )
  82                         );
  83                 } else {
  84                         $this->assertEquals( "<$tag></$tag>\n",
  85                                 Sanitizer::removeHTMLtags( "<$tag>" )
  86                         );
  87                 }
  88         }
  89
  90         /**
  91          * Provide HTML5 tags
  92          */
  93         function provideHtml5Tags() {
  94                 $ESCAPED  = true; # We want tag to be escaped
  95                 $VERBATIM = false;  # We want to keep the tag
  96                 return array(
  97                         array( 'data', $VERBATIM ),
  98                         array( 'mark', $VERBATIM ),
  99                         array( 'time', $VERBATIM ),
 100                         array( 'video', $ESCAPED ),
 101                 );
 102         }
 103
 104         function testSelfClosingTag() {
 105                 $this->setMwGlobals( array(
 106                         'wgUseTidy' => false
 107                 ));
 108
 109                 $this->assertEquals(
 110                         '<div>Hello world</div>',
 111                         Sanitizer::removeHTMLtags( '<div>Hello world</div />' ),
 112                         'Self-closing closing div'
 113                 );
 114         }
 115
 116         function testDecodeTagAttributes() {
 117                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo=bar' ), array( 'foo' => 'bar' ), 'Unquoted attribute' );
 118                 $this->assertEquals( Sanitizer::decodeTagAttributes( '    foo   =   bar    ' ), array( 'foo' => 'bar' ), 'Spaced attribute' );
 119                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo="bar"' ), array( 'foo' => 'bar' ), 'Double-quoted attribute' );
 120                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo=\'bar\'' ), array( 'foo' => 'bar' ), 'Single-quoted attribute' );
 121                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo=\'bar\'   baz="foo"' ), array( 'foo' => 'bar', 'baz' => 'foo' ), 'Several attributes' );
 122
 123                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo=\'bar\'   baz="foo"' ), array( 'foo' => 'bar', 'baz' => 'foo' ), 'Several attributes' );
 124                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo=\'bar\'   baz="foo"' ), array( 'foo' => 'bar', 'baz' => 'foo' ), 'Several attributes' );
 125
 126                 $this->assertEquals( Sanitizer::decodeTagAttributes( ':foo=\'bar\'' ), array( ':foo' => 'bar' ), 'Leading :' );
 127                 $this->assertEquals( Sanitizer::decodeTagAttributes( '_foo=\'bar\'' ), array( '_foo' => 'bar' ), 'Leading _' );
 128                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'Foo=\'bar\'' ), array( 'foo' => 'bar' ), 'Leading capital' );
 129                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'FOO=BAR' ), array( 'foo' => 'BAR' ), 'Attribute keys are normalized to lowercase' );
 130
 131                 # Invalid beginning
 132                 $this->assertEquals( Sanitizer::decodeTagAttributes( '-foo=bar' ), array(), 'Leading - is forbidden' );
 133                 $this->assertEquals( Sanitizer::decodeTagAttributes( '.foo=bar' ), array(), 'Leading . is forbidden' );
 134                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo-bar=bar' ), array( 'foo-bar' => 'bar' ), 'A - is allowed inside the attribute' );
 135                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo-=bar' ), array( 'foo-' => 'bar' ), 'A - is allowed inside the attribute' );
 136
 137                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo.bar=baz' ), array( 'foo.bar' => 'baz' ), 'A . is allowed inside the attribute' );
 138                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo.=baz' ), array( 'foo.' => 'baz' ), 'A . is allowed as last character' );
 139
 140                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo6=baz' ), array( 'foo6' => 'baz' ), 'Numbers are allowed' );
 141
 142                 # This bit is more relaxed than XML rules, but some extensions use it, like ProofreadPage (see bug 27539)
 143                 $this->assertEquals( Sanitizer::decodeTagAttributes( '1foo=baz' ), array( '1foo' => 'baz' ), 'Leading numbers are allowed' );
 144
 145                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo$=baz' ), array(), 'Symbols are not allowed' );
 146                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo@=baz' ), array(), 'Symbols are not allowed' );
 147                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo~=baz' ), array(), 'Symbols are not allowed' );
 148
 149
 150                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo=1[#^`*%w/(' ), array( 'foo' => '1[#^`*%w/(' ), 'All kind of characters are allowed as values' );
 151                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo="1[#^`*%\'w/("' ), array( 'foo' => '1[#^`*%\'w/(' ), 'Double quotes are allowed if quoted by single quotes' );
 152                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo=\'1[#^`*%"w/(\'' ), array( 'foo' => '1[#^`*%"w/(' ), 'Single quotes are allowed if quoted by double quotes' );
 153                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo=&amp;&quot;' ), array( 'foo' => '&"' ), 'Special chars can be provided as entities' );
 154                 $this->assertEquals( Sanitizer::decodeTagAttributes( 'foo=&foobar;' ), array( 'foo' => '&foobar;' ), 'Entity-like items are accepted' );
 155         }
 156
 157         /**
 158          * @dataProvider provideDeprecatedAttributes
 159          */
 160         function testDeprecatedAttributesUnaltered( $inputAttr, $inputEl ) {
 161
 162                 $this->assertEquals( " $inputAttr", Sanitizer::fixTagAttributes( $inputAttr, $inputEl ) );
 163         }
 164
 165         public static function provideDeprecatedAttributes() {
 166                 return array(
 167                         array( 'clear="left"', 'br' ),
 168                         array( 'clear="all"', 'br' ),
 169                         array( 'width="100"', 'td' ),
 170                         array( 'nowrap="true"', 'td' ),
 171                         array( 'nowrap=""', 'td' ),
 172                         array( 'align="right"', 'td' ),
 173                         array( 'align="center"', 'table' ),
 174                         array( 'align="left"', 'tr' ),
 175                         array( 'align="center"', 'div' ),
 176                         array( 'align="left"', 'h1' ),
 177                         array( 'align="left"', 'span' ),
 178                 );
 179         }
 180
 181         /**
 182          * @dataProvider provideCssCommentsFixtures
 183          */
 184         function testCssCommentsChecking( $expected, $css, $message = '' ) {
 185                 $this->assertEquals(
 186                         $expected,
 187                         Sanitizer::checkCss( $css ),
 188                         $message
 189                 );
 190         }
 191
 192         public static function provideCssCommentsFixtures() {
 193                 /** array( <expected>, <css>, [message] ) */
 194                 return array(
 195                         array( ' ', '/**/' ),
 196                         array( ' ', '/****/' ),
 197                         array( ' ', '/* comment */' ),
 198                         array( ' ', "\\2f\\2a foo \\2a\\2f",
 199                                 'Backslash-escaped comments must be stripped (bug 28450)' ),
 200                         array( '', '/* unfinished comment structure',
 201                                 'Remove anything after a comment-start token' ),
 202                         array( '', "\\2f\\2a unifinished comment'",
 203                                 'Remove anything after a backslash-escaped comment-start token' ),
 204                         array( '/* insecure input */', 'filter: progid:DXImageTransform.Microsoft.AlphaImageLoader(src=\'asdf.png\',sizingMethod=\'scale\');'),
 205                         array( '/* insecure input */', '-ms-filter: "progid:DXImageTransform.Microsoft.AlphaImageLoader(src=\'asdf.png\',sizingMethod=\'scale\')";'),
 206                         array( '/* insecure input */', 'width: expression(1+1);'),
 207                         array( '/* insecure input */', 'background-image: image(asdf.png);'),
 208                         array( '/* insecure input */', 'background-image: -webkit-image(asdf.png);'),
 209                         array( '/* insecure input */', 'background-image: -moz-image(asdf.png);'),
 210                         array( '/* insecure input */', 'background-image: image-set("asdf.png" 1x, "asdf.png" 2x);'),
 211                         array( '/* insecure input */', 'background-image: -webkit-image-set("asdf.png" 1x, "asdf.png" 2x);'),
 212                         array( '/* insecure input */', 'background-image: -moz-image-set("asdf.png" 1x, "asdf.png" 2x);'),
 213                 );
 214         }
 215 }
 216