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  * @author Daniel Kinzler
  20  */
  21
  22 use MediaWiki\Interwiki\InterwikiLookup;
  23
  24 /**
  25  * @covers MediaWikiTitleCodec
  26  *
  27  * @group Title
  28  * @group Database
  29  *        ^--- needed because of global state in
  30  */
  31 class MediaWikiTitleCodecTest extends MediaWikiTestCase {
  32
  33         public function setUp() {
  34                 parent::setUp();
  35
  36                 $this->setMwGlobals( [
  37                         'wgAllowUserJs' => false,
  38                         'wgDefaultLanguageVariant' => false,
  39                         'wgMetaNamespace' => 'Project',
  40                         'wgLocalInterwikis' => [ 'localtestiw' ],
  41                         'wgCapitalLinks' => true,
  42                 ] );
  43                 $this->setUserLang( 'en' );
  44                 $this->setContentLang( 'en' );
  45         }
  46
  47         /**
  48          * Returns a mock GenderCache that will consider a user "female" if the
  49          * first part of the user name ends with "a".
  50          *
  51          * @return GenderCache
  52          */
  53         private function getGenderCache() {
  54                 $genderCache = $this->getMockBuilder( GenderCache::class )
  55                         ->disableOriginalConstructor()
  56                         ->getMock();
  57
  58                 $genderCache->expects( $this->any() )
  59                         ->method( 'getGenderOf' )
  60                         ->will( $this->returnCallback( function ( $userName ) {
  61                                 return preg_match( '/^[^- _]+a( |_|$)/u', $userName ) ? 'female' : 'male';
  62                         } ) );
  63
  64                 return $genderCache;
  65         }
  66
  67         /**
  68          * Returns a mock InterwikiLookup that only has an isValidInterwiki() method, which recognizes
  69          * 'localtestiw' and 'remotetestiw'. All other methods throw.
  70          *
  71          * @return InterwikiLookup
  72          */
  73         private function getInterwikiLookup() : InterwikiLookup {
  74                 $iwLookup = $this->createMock( InterwikiLookup::class );
  75
  76                 $iwLookup->expects( $this->any() )
  77                         ->method( 'isValidInterwiki' )
  78                         ->will( $this->returnCallback( function ( $prefix ) {
  79                                 return $prefix === 'localtestiw' || $prefix === 'remotetestiw';
  80                         } ) );
  81
  82                 $iwLookup->expects( $this->never() )
  83                         ->method( $this->callback( function ( $name ) {
  84                                 return $name !== 'isValidInterwiki';
  85                         } ) );
  86
  87                 return $iwLookup;
  88         }
  89
  90         /**
  91          * Returns a mock NamespaceInfo that has only a hasGenderDistinction() method, which assumes
  92          * only NS_USER and NS_USER_TALK have a gender distinction. All other methods throw.
  93          *
  94          * @return NamespaceInfo
  95          */
  96         private function getNamespaceInfo() : NamespaceInfo {
  97                 $nsInfo = $this->createMock( NamespaceInfo::class );
  98
  99                 $nsInfo->expects( $this->any() )
 100                         ->method( 'hasGenderDistinction' )
 101                         ->will( $this->returnCallback( function ( $ns ) {
 102                                 return $ns === NS_USER || $ns === NS_USER_TALK;
 103                         } ) );
 104
 105                 $nsInfo->expects( $this->never() )
 106                         ->method( $this->callback( function ( $name ) {
 107                                 return $name !== 'hasGenderDistinction';
 108                         } ) );
 109
 110                 return $nsInfo;
 111         }
 112
 113         protected function makeCodec( $lang ) {
 114                 return new MediaWikiTitleCodec(
 115                         Language::factory( $lang ),
 116                         $this->getGenderCache(),
 117                         [],
 118                         $this->getInterwikiLookup(),
 119                         $this->getNamespaceInfo()
 120                 );
 121         }
 122
 123         public static function provideFormat() {
 124                 return [
 125                         [ NS_MAIN, 'Foo_Bar', '', '', 'en', 'Foo Bar' ],
 126                         [ NS_USER, 'Hansi_Maier', 'stuff_and_so_on', '', 'en', 'User:Hansi Maier#stuff and so on' ],
 127                         [ false, 'Hansi_Maier', '', '', 'en', 'Hansi Maier' ],
 128                         [
 129                                 NS_USER_TALK,
 130                                 'hansi__maier',
 131                                 '',
 132                                 '',
 133                                 'en',
 134                                 'User talk:hansi  maier',
 135                                 'User talk:Hansi maier'
 136                         ],
 137
 138                         // getGenderCache() provides a mock that considers first
 139                         // names ending in "a" to be female.
 140                         [ NS_USER, 'Lisa_Müller', '', '', 'de', 'Benutzerin:Lisa Müller' ],
 141                         [ NS_MAIN, 'FooBar', '', 'remotetestiw', 'en', 'remotetestiw:FooBar' ],
 142                 ];
 143         }
 144
 145         /**
 146          * @dataProvider provideFormat
 147          */
 148         public function testFormat( $namespace, $text, $fragment, $interwiki, $lang, $expected,
 149                 $normalized = null
 150         ) {
 151                 if ( $normalized === null ) {
 152                         $normalized = $expected;
 153                 }
 154
 155                 $codec = $this->makeCodec( $lang );
 156                 $actual = $codec->formatTitle( $namespace, $text, $fragment, $interwiki );
 157
 158                 $this->assertEquals( $expected, $actual, 'formatted' );
 159
 160                 // test round trip
 161                 $parsed = $codec->parseTitle( $actual, NS_MAIN );
 162                 $actual2 = $codec->formatTitle(
 163                         $parsed->getNamespace(),
 164                         $parsed->getText(),
 165                         $parsed->getFragment(),
 166                         $parsed->getInterwiki()
 167                 );
 168
 169                 $this->assertEquals( $normalized, $actual2, 'normalized after round trip' );
 170         }
 171
 172         public static function provideGetText() {
 173                 return [
 174                         [ NS_MAIN, 'Foo_Bar', '', 'en', 'Foo Bar' ],
 175                         [ NS_USER, 'Hansi_Maier', 'stuff_and_so_on', 'en', 'Hansi Maier' ],
 176                 ];
 177         }
 178
 179         /**
 180          * @dataProvider provideGetText
 181          */
 182         public function testGetText( $namespace, $dbkey, $fragment, $lang, $expected ) {
 183                 $codec = $this->makeCodec( $lang );
 184                 $title = new TitleValue( $namespace, $dbkey, $fragment );
 185
 186                 $actual = $codec->getText( $title );
 187
 188                 $this->assertEquals( $expected, $actual );
 189         }
 190
 191         public static function provideGetPrefixedText() {
 192                 return [
 193                         [ NS_MAIN, 'Foo_Bar', '', 'en', 'Foo Bar' ],
 194                         [ NS_USER, 'Hansi_Maier', 'stuff_and_so_on', 'en', 'User:Hansi Maier' ],
 195
 196                         // No capitalization or normalization is applied while formatting!
 197                         [ NS_USER_TALK, 'hansi__maier', '', 'en', 'User talk:hansi  maier' ],
 198
 199                         // getGenderCache() provides a mock that considers first
 200                         // names ending in "a" to be female.
 201                         [ NS_USER, 'Lisa_Müller', '', 'de', 'Benutzerin:Lisa Müller' ],
 202                         [ 1000000, 'Invalid_namespace', '', 'en', 'Special:Badtitle/NS1000000:Invalid namespace' ],
 203                 ];
 204         }
 205
 206         /**
 207          * @dataProvider provideGetPrefixedText
 208          */
 209         public function testGetPrefixedText( $namespace, $dbkey, $fragment, $lang, $expected ) {
 210                 $codec = $this->makeCodec( $lang );
 211                 $title = new TitleValue( $namespace, $dbkey, $fragment );
 212
 213                 $actual = $codec->getPrefixedText( $title );
 214
 215                 $this->assertEquals( $expected, $actual );
 216         }
 217
 218         public static function provideGetPrefixedDBkey() {
 219                 return [
 220                         [ NS_MAIN, 'Foo_Bar', '', '', 'en', 'Foo_Bar' ],
 221                         [ NS_USER, 'Hansi_Maier', 'stuff_and_so_on', '', 'en', 'User:Hansi_Maier' ],
 222
 223                         // No capitalization or normalization is applied while formatting!
 224                         [ NS_USER_TALK, 'hansi__maier', '', '', 'en', 'User_talk:hansi__maier' ],
 225
 226                         // getGenderCache() provides a mock that considers first
 227                         // names ending in "a" to be female.
 228                         [ NS_USER, 'Lisa_Müller', '', '', 'de', 'Benutzerin:Lisa_Müller' ],
 229
 230                         [ NS_MAIN, 'Remote_page', '', 'remotetestiw', 'en', 'remotetestiw:Remote_page' ],
 231
 232                         // non-existent namespace
 233                         [ 10000000, 'Foobar', '', '', 'en', 'Special:Badtitle/NS10000000:Foobar' ],
 234                 ];
 235         }
 236
 237         /**
 238          * @dataProvider provideGetPrefixedDBkey
 239          */
 240         public function testGetPrefixedDBkey( $namespace, $dbkey, $fragment,
 241                 $interwiki, $lang, $expected
 242         ) {
 243                 $codec = $this->makeCodec( $lang );
 244                 $title = new TitleValue( $namespace, $dbkey, $fragment, $interwiki );
 245
 246                 $actual = $codec->getPrefixedDBkey( $title );
 247
 248                 $this->assertEquals( $expected, $actual );
 249         }
 250
 251         public static function provideGetFullText() {
 252                 return [
 253                         [ NS_MAIN, 'Foo_Bar', '', 'en', 'Foo Bar' ],
 254                         [ NS_USER, 'Hansi_Maier', 'stuff_and_so_on', 'en', 'User:Hansi Maier#stuff and so on' ],
 255
 256                         // No capitalization or normalization is applied while formatting!
 257                         [ NS_USER_TALK, 'hansi__maier', '', 'en', 'User talk:hansi  maier' ],
 258                 ];
 259         }
 260
 261         /**
 262          * @dataProvider provideGetFullText
 263          */
 264         public function testGetFullText( $namespace, $dbkey, $fragment, $lang, $expected ) {
 265                 $codec = $this->makeCodec( $lang );
 266                 $title = new TitleValue( $namespace, $dbkey, $fragment );
 267
 268                 $actual = $codec->getFullText( $title );
 269
 270                 $this->assertEquals( $expected, $actual );
 271         }
 272
 273         public static function provideParseTitle() {
 274                 // TODO: test capitalization and trimming
 275                 // TODO: test unicode normalization
 276
 277                 return [
 278                         [ '  : Hansi_Maier _ ', NS_MAIN, 'en',
 279                                 new TitleValue( NS_MAIN, 'Hansi_Maier', '' ) ],
 280                         [ 'User:::1', NS_MAIN, 'de',
 281                                 new TitleValue( NS_USER, '0:0:0:0:0:0:0:1', '' ) ],
 282                         [ ' lisa Müller', NS_USER, 'de',
 283                                 new TitleValue( NS_USER, 'Lisa_Müller', '' ) ],
 284                         [ 'benutzerin:lisa Müller#stuff', NS_MAIN, 'de',
 285                                 new TitleValue( NS_USER, 'Lisa_Müller', 'stuff' ) ],
 286
 287                         [ ':Category:Quux', NS_MAIN, 'en',
 288                                 new TitleValue( NS_CATEGORY, 'Quux', '' ) ],
 289                         [ 'Category:Quux', NS_MAIN, 'en',
 290                                 new TitleValue( NS_CATEGORY, 'Quux', '' ) ],
 291                         [ 'Category:Quux', NS_CATEGORY, 'en',
 292                                 new TitleValue( NS_CATEGORY, 'Quux', '' ) ],
 293                         [ 'Quux', NS_CATEGORY, 'en',
 294                                 new TitleValue( NS_CATEGORY, 'Quux', '' ) ],
 295                         [ ':Quux', NS_CATEGORY, 'en',
 296                                 new TitleValue( NS_MAIN, 'Quux', '' ) ],
 297
 298                         // getGenderCache() provides a mock that considers first
 299                         // names ending in "a" to be female.
 300
 301                         [ 'a b c', NS_MAIN, 'en',
 302                                 new TitleValue( NS_MAIN, 'A_b_c' ) ],
 303                         [ ' a  b  c ', NS_MAIN, 'en',
 304                                 new TitleValue( NS_MAIN, 'A_b_c' ) ],
 305                         [ ' _ Foo __ Bar_ _', NS_MAIN, 'en',
 306                                 new TitleValue( NS_MAIN, 'Foo_Bar' ) ],
 307
 308                         // NOTE: cases copied from TitleTest::testSecureAndSplit. Keep in sync.
 309                         [ 'Sandbox', NS_MAIN, 'en', ],
 310                         [ 'A "B"', NS_MAIN, 'en', ],
 311                         [ 'A \'B\'', NS_MAIN, 'en', ],
 312                         [ '.com', NS_MAIN, 'en', ],
 313                         [ '~', NS_MAIN, 'en', ],
 314                         [ '"', NS_MAIN, 'en', ],
 315                         [ '\'', NS_MAIN, 'en', ],
 316
 317                         [ 'Talk:Sandbox', NS_MAIN, 'en',
 318                                 new TitleValue( NS_TALK, 'Sandbox' ) ],
 319                         [ 'Talk:Foo:Sandbox', NS_MAIN, 'en',
 320                                 new TitleValue( NS_TALK, 'Foo:Sandbox' ) ],
 321                         [ 'File:Example.svg', NS_MAIN, 'en',
 322                                 new TitleValue( NS_FILE, 'Example.svg' ) ],
 323                         [ 'File_talk:Example.svg', NS_MAIN, 'en',
 324                                 new TitleValue( NS_FILE_TALK, 'Example.svg' ) ],
 325                         [ 'Foo/.../Sandbox', NS_MAIN, 'en',
 326                                 'Foo/.../Sandbox' ],
 327                         [ 'Sandbox/...', NS_MAIN, 'en',
 328                                 'Sandbox/...' ],
 329                         [ 'A~~', NS_MAIN, 'en',
 330                                 'A~~' ],
 331                         // Length is 256 total, but only title part matters
 332                         [ 'Category:' . str_repeat( 'x', 248 ), NS_MAIN, 'en',
 333                                 new TitleValue( NS_CATEGORY,
 334                                         'X' . str_repeat( 'x', 247 ) ) ],
 335                         [ str_repeat( 'x', 252 ), NS_MAIN, 'en',
 336                                 'X' . str_repeat( 'x', 251 ) ],
 337                         // Test decoding and normalization
 338                         [ '&quot;n&#x303;&#34;', NS_MAIN, 'en', new TitleValue( NS_MAIN, '"ñ"' ) ],
 339                         [ 'X#n&#x303;', NS_MAIN, 'en', new TitleValue( NS_MAIN, 'X', 'ñ' ) ],
 340                         // target section parsing
 341                         'empty fragment' => [ 'X#', NS_MAIN, 'en', new TitleValue( NS_MAIN, 'X' ) ],
 342                         'double hash' => [ 'X##', NS_MAIN, 'en', new TitleValue( NS_MAIN, 'X', '#' ) ],
 343                         'fragment with hash' => [ 'X#z#z', NS_MAIN, 'en', new TitleValue( NS_MAIN, 'X', 'z#z' ) ],
 344                         'fragment with space' => [ 'X#z z', NS_MAIN, 'en', new TitleValue( NS_MAIN, 'X', 'z z' ) ],
 345                         'fragment with percent' => [ 'X#z%z', NS_MAIN, 'en', new TitleValue( NS_MAIN, 'X', 'z%z' ) ],
 346                         'fragment with amp' => [ 'X#z&z', NS_MAIN, 'en', new TitleValue( NS_MAIN, 'X', 'z&z' ) ],
 347                 ];
 348         }
 349
 350         /**
 351          * @dataProvider provideParseTitle
 352          */
 353         public function testParseTitle( $text, $ns, $lang, $title = null ) {
 354                 if ( $title === null ) {
 355                         $title = str_replace( ' ', '_', trim( $text ) );
 356                 }
 357
 358                 if ( is_string( $title ) ) {
 359                         $title = new TitleValue( NS_MAIN, $title, '' );
 360                 }
 361
 362                 $codec = $this->makeCodec( $lang );
 363                 $actual = $codec->parseTitle( $text, $ns );
 364
 365                 $this->assertEquals( $title, $actual );
 366         }
 367
 368         public static function provideParseTitle_invalid() {
 369                 // TODO: test unicode errors
 370
 371                 return [
 372                         [ '#' ],
 373                         [ '::' ],
 374                         [ '::xx' ],
 375                         [ '::##' ],
 376                         [ ' :: x' ],
 377
 378                         [ 'Talk:File:Foo.jpg' ],
 379                         [ 'Talk:localtestiw:Foo' ],
 380                         [ '::1' ], // only valid in user namespace
 381                         [ 'User::x' ], // leading ":" in a user name is only valid of IPv6 addresses
 382
 383                         // NOTE: cases copied from TitleTest::testSecureAndSplit. Keep in sync.
 384                         [ '' ],
 385                         [ ':' ],
 386                         [ '__  __' ],
 387                         [ '  __  ' ],
 388                         // Bad characters forbidden regardless of wgLegalTitleChars
 389                         [ 'A [ B' ],
 390                         [ 'A ] B' ],
 391                         [ 'A { B' ],
 392                         [ 'A } B' ],
 393                         [ 'A < B' ],
 394                         [ 'A > B' ],
 395                         [ 'A | B' ],
 396                         // URL encoding
 397                         [ 'A%20B' ],
 398                         [ 'A%23B' ],
 399                         [ 'A%2523B' ],
 400                         // XML/HTML character entity references
 401                         // Note: Commented out because they are not marked invalid by the PHP test as
 402                         // Title::newFromText runs Sanitizer::decodeCharReferencesAndNormalize first.
 403                         // [ 'A &eacute; B' ],
 404                         // [ 'A &#233; B' ],
 405                         // [ 'A &#x00E9; B' ],
 406                         // Subject of NS_TALK does not roundtrip to NS_MAIN
 407                         [ 'Talk:File:Example.svg' ],
 408                         // Directory navigation
 409                         [ '.' ],
 410                         [ '..' ],
 411                         [ './Sandbox' ],
 412                         [ '../Sandbox' ],
 413                         [ 'Foo/./Sandbox' ],
 414                         [ 'Foo/../Sandbox' ],
 415                         [ 'Sandbox/.' ],
 416                         [ 'Sandbox/..' ],
 417                         // Tilde
 418                         [ 'A ~~~ Name' ],
 419                         [ 'A ~~~~ Signature' ],
 420                         [ 'A ~~~~~ Timestamp' ],
 421                         [ str_repeat( 'x', 256 ) ],
 422                         // Namespace prefix without actual title
 423                         [ 'Talk:' ],
 424                         [ 'Category: ' ],
 425                         [ 'Category: #bar' ]
 426                 ];
 427         }
 428
 429         /**
 430          * @dataProvider provideParseTitle_invalid
 431          */
 432         public function testParseTitle_invalid( $text ) {
 433                 $this->setExpectedException( MalformedTitleException::class );
 434
 435                 $codec = $this->makeCodec( 'en' );
 436                 $codec->parseTitle( $text, NS_MAIN );
 437         }
 438
 439         public static function provideGetNamespaceName() {
 440                 return [
 441                         [ NS_MAIN, 'Foo', 'en', '' ],
 442                         [ NS_USER, 'Foo', 'en', 'User' ],
 443                         [ NS_USER, 'Hansi Maier', 'de', 'Benutzer' ],
 444
 445                         // getGenderCache() provides a mock that considers first
 446                         // names ending in "a" to be female.
 447                         [ NS_USER, 'Lisa Müller', 'de', 'Benutzerin' ],
 448                 ];
 449         }
 450
 451         /**
 452          * @dataProvider provideGetNamespaceName
 453          */
 454         public function testGetNamespaceName( $namespace, $text, $lang, $expected ) {
 455                 $codec = $this->makeCodec( $lang );
 456                 $name = $codec->getNamespaceName( $namespace, $text );
 457
 458                 $this->assertEquals( $expected, $name );
 459         }
 460 }