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