tests/phpunit/includes/title/MediaWikiTitleCodecTest.php

   1 <?php
   2 /**
   3  * This program is free software; you can redistribute it and/or modify
   4  * it under the terms of the GNU General Public License as published by
   5  * the Free Software Foundation; either version 2 of the License, or
   6  * (at your option) any later version.
   7  *
   8  * This program is distributed in the hope that it will be useful,
   9  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  10  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  11  * GNU General Public License for more details.
  12  *
  13  * You should have received a copy of the GNU General Public License along
  14  * with this program; if not, write to the Free Software Foundation, Inc.,
  15  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  16  * http://www.gnu.org/copyleft/gpl.html
  17  *
  18  * @file
  19  * @license GPL 2+
  20  * @author Daniel Kinzler
  21  */
  22
  23 /**
  24  * @covers MediaWikiTitleCodec
  25  *
  26  * @group Title
  27  * @group Database
  28  *        ^--- needed because of global state in
  29  */
  30 class MediaWikiTitleCodecTest extends MediaWikiTestCase {
  31
  32         public function setUp() {
  33                 parent::setUp();
  34
  35                 $this->setMwGlobals( array(
  36                         'wgLanguageCode' => 'en',
  37                         'wgContLang' => Language::factory( 'en' ),
  38                         // User language
  39                         'wgLang' => Language::factory( 'en' ),
  40                         'wgAllowUserJs' => false,
  41                         'wgDefaultLanguageVariant' => false,
  42                         'wgLocalInterwikis' => array( 'localtestiw' ),
  43
  44                         // NOTE: this is why global state is evil.
  45                         // TODO: refactor access to the interwiki codes so it can be injected.
  46                         'wgHooks' => array(
  47                                 'InterwikiLoadPrefix' => array(
  48                                         function ( $prefix, &$data ) {
  49                                                 if ( $prefix === 'localtestiw' ) {
  50                                                         $data = array( 'iw_url' => 'localtestiw' );
  51                                                 } elseif ( $prefix === 'remotetestiw' ) {
  52                                                         $data = array( 'iw_url' => 'remotetestiw' );
  53                                                 }
  54                                                 return false;
  55                                         }
  56                                 )
  57                         )
  58                 ) );
  59         }
  60
  61         /**
  62          * Returns a mock GenderCache that will consider a user "female" if the
  63          * first part of the user name ends with "a".
  64          *
  65          * @return GenderCache
  66          */
  67         private function getGenderCache() {
  68                 $genderCache = $this->getMockBuilder( 'GenderCache' )
  69                         ->disableOriginalConstructor()
  70                         ->getMock();
  71
  72                 $genderCache->expects( $this->any() )
  73                         ->method( 'getGenderOf' )
  74                         ->will( $this->returnCallback( function( $userName ) {
  75                                 return preg_match( '/^[^- _]+a( |_|$)/u', $userName ) ? 'female' : 'male';
  76                         } ) );
  77
  78                 return $genderCache;
  79         }
  80
  81         protected function makeCodec( $lang ) {
  82                 $gender = $this->getGenderCache();
  83                 $lang = Language::factory( $lang );
  84                 return new MediaWikiTitleCodec( $lang, $gender );
  85         }
  86
  87         public function provideFormat() {
  88                 return array(
  89                         array( NS_MAIN, 'Foo_Bar', '', 'en', 'Foo Bar' ),
  90                         array( NS_USER, 'Hansi_Maier', 'stuff_and_so_on', 'en', 'User:Hansi Maier#stuff and so on' ),
  91                         array( false, 'Hansi_Maier', '', 'en', 'Hansi Maier' ),
  92                         array(
  93                                 NS_USER_TALK,
  94                                 'hansi__maier',
  95                                 '',
  96                                 'en',
  97                                 'User talk:hansi  maier',
  98                                 'User talk:Hansi maier'
  99                         ),
 100
 101                         // getGenderCache() provides a mock that considers first
 102                         // names ending in "a" to be female.
 103                         array( NS_USER, 'Lisa_Müller', '', 'de', 'Benutzerin:Lisa Müller' ),
 104                 );
 105         }
 106
 107         /**
 108          * @dataProvider provideFormat
 109          */
 110         public function testFormat( $namespace, $text, $fragment, $lang, $expected, $normalized = null ) {
 111                 if ( $normalized === null ) {
 112                         $normalized = $expected;
 113                 }
 114
 115                 $codec = $this->makeCodec( $lang );
 116                 $actual = $codec->formatTitle( $namespace, $text, $fragment );
 117
 118                 $this->assertEquals( $expected, $actual, 'formatted' );
 119
 120                 // test round trip
 121                 $parsed = $codec->parseTitle( $actual, NS_MAIN );
 122                 $actual2 = $codec->formatTitle(
 123                         $parsed->getNamespace(),
 124                         $parsed->getText(),
 125                         $parsed->getFragment()
 126                 );
 127
 128                 $this->assertEquals( $normalized, $actual2, 'normalized after round trip' );
 129         }
 130
 131         public function provideGetText() {
 132                 return array(
 133                         array( NS_MAIN, 'Foo_Bar', '', 'en', 'Foo Bar' ),
 134                         array( NS_USER, 'Hansi_Maier', 'stuff_and_so_on', 'en', 'Hansi Maier' ),
 135                 );
 136         }
 137
 138         /**
 139          * @dataProvider provideGetText
 140          */
 141         public function testGetText( $namespace, $dbkey, $fragment, $lang, $expected ) {
 142                 $codec = $this->makeCodec( $lang );
 143                 $title = new TitleValue( $namespace, $dbkey, $fragment );
 144
 145                 $actual = $codec->getText( $title );
 146
 147                 $this->assertEquals( $expected, $actual );
 148         }
 149
 150         public function provideGetPrefixedText() {
 151                 return array(
 152                         array( NS_MAIN, 'Foo_Bar', '', 'en', 'Foo Bar' ),
 153                         array( NS_USER, 'Hansi_Maier', 'stuff_and_so_on', 'en', 'User:Hansi Maier' ),
 154
 155                         // No capitalization or normalization is applied while formatting!
 156                         array( NS_USER_TALK, 'hansi__maier', '', 'en', 'User talk:hansi  maier' ),
 157
 158                         // getGenderCache() provides a mock that considers first
 159                         // names ending in "a" to be female.
 160                         array( NS_USER, 'Lisa_Müller', '', 'de', 'Benutzerin:Lisa Müller' ),
 161                 );
 162         }
 163
 164         /**
 165          * @dataProvider provideGetPrefixedText
 166          */
 167         public function testGetPrefixedText( $namespace, $dbkey, $fragment, $lang, $expected ) {
 168                 $codec = $this->makeCodec( $lang );
 169                 $title = new TitleValue( $namespace, $dbkey, $fragment );
 170
 171                 $actual = $codec->getPrefixedText( $title );
 172
 173                 $this->assertEquals( $expected, $actual );
 174         }
 175
 176         public function provideGetFullText() {
 177                 return array(
 178                         array( NS_MAIN, 'Foo_Bar', '', 'en', 'Foo Bar' ),
 179                         array( NS_USER, 'Hansi_Maier', 'stuff_and_so_on', 'en', 'User:Hansi Maier#stuff and so on' ),
 180
 181                         // No capitalization or normalization is applied while formatting!
 182                         array( NS_USER_TALK, 'hansi__maier', '', 'en', 'User talk:hansi  maier' ),
 183                 );
 184         }
 185
 186         /**
 187          * @dataProvider provideGetFullText
 188          */
 189         public function testGetFullText( $namespace, $dbkey, $fragment, $lang, $expected ) {
 190                 $codec = $this->makeCodec( $lang );
 191                 $title = new TitleValue( $namespace, $dbkey, $fragment );
 192
 193                 $actual = $codec->getFullText( $title );
 194
 195                 $this->assertEquals( $expected, $actual );
 196         }
 197
 198         public function provideParseTitle() {
 199                 //TODO: test capitalization and trimming
 200                 //TODO: test unicode normalization
 201
 202                 return array(
 203                         array( '  : Hansi_Maier _ ', NS_MAIN, 'en',
 204                                 new TitleValue( NS_MAIN, 'Hansi_Maier', '' ) ),
 205                         array( 'User:::1', NS_MAIN, 'de',
 206                                 new TitleValue( NS_USER, '0:0:0:0:0:0:0:1', '' ) ),
 207                         array( ' lisa Müller', NS_USER, 'de',
 208                                 new TitleValue( NS_USER, 'Lisa_Müller', '' ) ),
 209                         array( 'benutzerin:lisa Müller#stuff', NS_MAIN, 'de',
 210                                 new TitleValue( NS_USER, 'Lisa_Müller', 'stuff' ) ),
 211
 212                         array( ':Category:Quux', NS_MAIN, 'en',
 213                                 new TitleValue( NS_CATEGORY, 'Quux', '' ) ),
 214                         array( 'Category:Quux', NS_MAIN, 'en',
 215                                 new TitleValue( NS_CATEGORY, 'Quux', '' ) ),
 216                         array( 'Category:Quux', NS_CATEGORY, 'en',
 217                                 new TitleValue( NS_CATEGORY, 'Quux', '' ) ),
 218                         array( 'Quux', NS_CATEGORY, 'en',
 219                                 new TitleValue( NS_CATEGORY, 'Quux', '' ) ),
 220                         array( ':Quux', NS_CATEGORY, 'en',
 221                                 new TitleValue( NS_MAIN, 'Quux', '' ) ),
 222
 223                         // getGenderCache() provides a mock that considers first
 224                         // names ending in "a" to be female.
 225
 226                         array( 'a b c', NS_MAIN, 'en',
 227                                 new TitleValue( NS_MAIN, 'A_b_c' ) ),
 228                         array( ' a  b  c ', NS_MAIN, 'en',
 229                                 new TitleValue( NS_MAIN, 'A_b_c' ) ),
 230                         array( ' _ Foo __ Bar_ _', NS_MAIN, 'en',
 231                                 new TitleValue( NS_MAIN, 'Foo_Bar' ) ),
 232
 233                         //NOTE: cases copied from TitleTest::testSecureAndSplit. Keep in sync.
 234                         array( 'Sandbox', NS_MAIN, 'en', ),
 235                         array( 'A "B"', NS_MAIN, 'en', ),
 236                         array( 'A \'B\'', NS_MAIN, 'en', ),
 237                         array( '.com', NS_MAIN, 'en', ),
 238                         array( '~', NS_MAIN, 'en', ),
 239                         array( '"', NS_MAIN, 'en', ),
 240                         array( '\'', NS_MAIN, 'en', ),
 241
 242                         array( 'Talk:Sandbox', NS_MAIN, 'en',
 243                                 new TitleValue( NS_TALK, 'Sandbox' ) ),
 244                         array( 'Talk:Foo:Sandbox', NS_MAIN, 'en',
 245                                 new TitleValue( NS_TALK, 'Foo:Sandbox' ) ),
 246                         array( 'File:Example.svg', NS_MAIN, 'en',
 247                                 new TitleValue( NS_FILE, 'Example.svg' ) ),
 248                         array( 'File_talk:Example.svg', NS_MAIN, 'en',
 249                                 new TitleValue( NS_FILE_TALK, 'Example.svg' ) ),
 250                         array( 'Foo/.../Sandbox', NS_MAIN, 'en',
 251                                 'Foo/.../Sandbox' ),
 252                         array( 'Sandbox/...', NS_MAIN, 'en',
 253                                 'Sandbox/...' ),
 254                         array( 'A~~', NS_MAIN, 'en',
 255                                 'A~~' ),
 256                         // Length is 256 total, but only title part matters
 257                         array( 'Category:' . str_repeat( 'x', 248 ), NS_MAIN, 'en',
 258                                 new TitleValue( NS_CATEGORY,
 259                                         'X' . str_repeat( 'x', 247 ) ) ),
 260                         array( str_repeat( 'x', 252 ), NS_MAIN, 'en',
 261                                 'X' . str_repeat( 'x', 251 ) )
 262                 );
 263         }
 264
 265         /**
 266          * @dataProvider provideParseTitle
 267          */
 268         public function testParseTitle( $text, $ns, $lang, $title = null ) {
 269                 if ( $title === null ) {
 270                         $title = str_replace( ' ', '_', trim( $text ) );
 271                 }
 272
 273                 if ( is_string( $title ) ) {
 274                         $title = new TitleValue( NS_MAIN, $title, '' );
 275                 }
 276
 277                 $codec = $this->makeCodec( $lang );
 278                 $actual = $codec->parseTitle( $text, $ns );
 279
 280                 $this->assertEquals( $title, $actual );
 281         }
 282
 283         public function provideParseTitle_invalid() {
 284                 //TODO: test unicode errors
 285
 286                 return array(
 287                         array( '#' ),
 288                         array( '::' ),
 289                         array( '::xx' ),
 290                         array( '::##' ),
 291                         array( ' :: x' ),
 292
 293                         array( 'Talk:File:Foo.jpg' ),
 294                         array( 'Talk:localtestiw:Foo' ),
 295                         array( 'remotetestiw:Foo' ),
 296                         array( '::1' ), // only valid in user namespace
 297                         array( 'User::x' ), // leading ":" in a user name is only valid of IPv6 addresses
 298
 299                         //NOTE: cases copied from TitleTest::testSecureAndSplit. Keep in sync.
 300                         array( '' ),
 301                         array( ':' ),
 302                         array( '__  __' ),
 303                         array( '  __  ' ),
 304                         // Bad characters forbidden regardless of wgLegalTitleChars
 305                         array( 'A [ B' ),
 306                         array( 'A ] B' ),
 307                         array( 'A { B' ),
 308                         array( 'A } B' ),
 309                         array( 'A < B' ),
 310                         array( 'A > B' ),
 311                         array( 'A | B' ),
 312                         // URL encoding
 313                         array( 'A%20B' ),
 314                         array( 'A%23B' ),
 315                         array( 'A%2523B' ),
 316                         // XML/HTML character entity references
 317                         // Note: Commented out because they are not marked invalid by the PHP test as
 318                         // Title::newFromText runs Sanitizer::decodeCharReferencesAndNormalize first.
 319                         //array( 'A &eacute; B' ),
 320                         //array( 'A &#233; B' ),
 321                         //array( 'A &#x00E9; B' ),
 322                         // Subject of NS_TALK does not roundtrip to NS_MAIN
 323                         array( 'Talk:File:Example.svg' ),
 324                         // Directory navigation
 325                         array( '.' ),
 326                         array( '..' ),
 327                         array( './Sandbox' ),
 328                         array( '../Sandbox' ),
 329                         array( 'Foo/./Sandbox' ),
 330                         array( 'Foo/../Sandbox' ),
 331                         array( 'Sandbox/.' ),
 332                         array( 'Sandbox/..' ),
 333                         // Tilde
 334                         array( 'A ~~~ Name' ),
 335                         array( 'A ~~~~ Signature' ),
 336                         array( 'A ~~~~~ Timestamp' ),
 337                         array( str_repeat( 'x', 256 ) ),
 338                         // Namespace prefix without actual title
 339                         array( 'Talk:' ),
 340                         array( 'Category: ' ),
 341                         array( 'Category: #bar' )
 342                 );
 343         }
 344
 345         /**
 346          * @dataProvider provideParseTitle_invalid
 347          */
 348         public function testParseTitle_invalid( $text ) {
 349                 $this->setExpectedException( 'MalformedTitleException' );
 350
 351                 $codec = $this->makeCodec( 'en' );
 352                 $codec->parseTitle( $text, NS_MAIN );
 353         }
 354
 355         public function provideGetNamespaceName() {
 356                 return array(
 357                         array( NS_MAIN, 'Foo', 'en', '' ),
 358                         array( NS_USER, 'Foo', 'en', 'User' ),
 359                         array( NS_USER, 'Hansi Maier', 'de', 'Benutzer' ),
 360
 361                         // getGenderCache() provides a mock that considers first
 362                         // names ending in "a" to be female.
 363                         array( NS_USER, 'Lisa Müller', 'de', 'Benutzerin' ),
 364                 );
 365         }
 366
 367         /**
 368          * @dataProvider provideGetNamespaceName
 369          *
 370          * @param int $namespace
 371          * @param string $text
 372          * @param string $lang
 373          * @param string $expected
 374          *
 375          * @internal param \TitleValue $title
 376          */
 377         public function testGetNamespaceName( $namespace, $text, $lang, $expected ) {
 378                 $codec = $this->makeCodec( $lang );
 379                 $name = $codec->getNamespaceName( $namespace, $text );
 380
 381                 $this->assertEquals( $expected, $name );
 382         }
 383 }