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                         // Strip soft hyphen and Unicode directional formatting characters
 108                         [ NS_MAIN, "Foo\xC2\xAD\xD8\x9C\xE2\x80\x8E\xE2\x80\x8F\xE2\x80\xAA\xE2\x80\xAB" .
 109                                 "\xE2\x80\xAC\xE2\x80\xAD\xE2\x80\xAE\xE2\x81\xA6\xE2\x81\xA7" .
 110                                 "\xE2\x81\xA8\xE2\x81\xA9bar", '', '', 'en',
 111                                 "Foo\xC2\xAD\xD8\x9C\xE2\x80\x8E\xE2\x80\x8F\xE2\x80\xAA\xE2\x80\xAB" .
 112                                 "\xE2\x80\xAC\xE2\x80\xAD\xE2\x80\xAE\xE2\x81\xA6\xE2\x81\xA7" .
 113                                 "\xE2\x81\xA8\xE2\x81\xA9bar", 'Foobar' ],
 114                 ];
 115         }
 116
 117         /**
 118          * @dataProvider provideFormat
 119          */
 120         public function testFormat( $namespace, $text, $fragment, $interwiki, $lang, $expected,
 121                 $normalized = null
 122         ) {
 123                 if ( $normalized === null ) {
 124                         $normalized = $expected;
 125                 }
 126
 127                 $codec = $this->makeCodec( $lang );
 128                 $actual = $codec->formatTitle( $namespace, $text, $fragment, $interwiki );
 129
 130                 $this->assertEquals( $expected, $actual, 'formatted' );
 131
 132                 // test round trip
 133                 $parsed = $codec->parseTitle( $actual, NS_MAIN );
 134                 $actual2 = $codec->formatTitle(
 135                         $parsed->getNamespace(),
 136                         $parsed->getText(),
 137                         $parsed->getFragment(),
 138                         $parsed->getInterwiki()
 139                 );
 140
 141                 $this->assertEquals( $normalized, $actual2, 'normalized after round trip' );
 142         }
 143
 144         public static function provideGetText() {
 145                 return [
 146                         [ NS_MAIN, 'Foo_Bar', '', 'en', 'Foo Bar' ],
 147                         [ NS_USER, 'Hansi_Maier', 'stuff_and_so_on', 'en', 'Hansi Maier' ],
 148                 ];
 149         }
 150
 151         /**
 152          * @dataProvider provideGetText
 153          */
 154         public function testGetText( $namespace, $dbkey, $fragment, $lang, $expected ) {
 155                 $codec = $this->makeCodec( $lang );
 156                 $title = new TitleValue( $namespace, $dbkey, $fragment );
 157
 158                 $actual = $codec->getText( $title );
 159
 160                 $this->assertEquals( $expected, $actual );
 161         }
 162
 163         public static function provideGetPrefixedText() {
 164                 return [
 165                         [ NS_MAIN, 'Foo_Bar', '', 'en', 'Foo Bar' ],
 166                         [ NS_USER, 'Hansi_Maier', 'stuff_and_so_on', 'en', 'User:Hansi Maier' ],
 167
 168                         // No capitalization or normalization is applied while formatting!
 169                         [ NS_USER_TALK, 'hansi__maier', '', 'en', 'User talk:hansi  maier' ],
 170
 171                         // getGenderCache() provides a mock that considers first
 172                         // names ending in "a" to be female.
 173                         [ NS_USER, 'Lisa_Müller', '', 'de', 'Benutzerin:Lisa Müller' ],
 174                         [ 1000000, 'Invalid_namespace', '', 'en', ':Invalid namespace' ],
 175                 ];
 176         }
 177
 178         /**
 179          * @dataProvider provideGetPrefixedText
 180          */
 181         public function testGetPrefixedText( $namespace, $dbkey, $fragment, $lang, $expected ) {
 182                 $codec = $this->makeCodec( $lang );
 183                 $title = new TitleValue( $namespace, $dbkey, $fragment );
 184
 185                 $actual = $codec->getPrefixedText( $title );
 186
 187                 $this->assertEquals( $expected, $actual );
 188         }
 189
 190         public static function provideGetPrefixedDBkey() {
 191                 return [
 192                         [ NS_MAIN, 'Foo_Bar', '', '', 'en', 'Foo_Bar' ],
 193                         [ NS_USER, 'Hansi_Maier', 'stuff_and_so_on', '', 'en', 'User:Hansi_Maier' ],
 194
 195                         // No capitalization or normalization is applied while formatting!
 196                         [ NS_USER_TALK, 'hansi__maier', '', '', 'en', 'User_talk:hansi__maier' ],
 197
 198                         // getGenderCache() provides a mock that considers first
 199                         // names ending in "a" to be female.
 200                         [ NS_USER, 'Lisa_Müller', '', '', 'de', 'Benutzerin:Lisa_Müller' ],
 201
 202                         [ NS_MAIN, 'Remote_page', '', 'remotetestiw', 'en', 'remotetestiw:Remote_page' ],
 203
 204                         // non-existent namespace
 205                         [ 10000000, 'Foobar', '', '', 'en', ':Foobar' ],
 206                 ];
 207         }
 208
 209         /**
 210          * @dataProvider provideGetPrefixedDBkey
 211          */
 212         public function testGetPrefixedDBkey( $namespace, $dbkey, $fragment,
 213                 $interwiki, $lang, $expected
 214         ) {
 215                 $codec = $this->makeCodec( $lang );
 216                 $title = new TitleValue( $namespace, $dbkey, $fragment, $interwiki );
 217
 218                 $actual = $codec->getPrefixedDBkey( $title );
 219
 220                 $this->assertEquals( $expected, $actual );
 221         }
 222
 223         public static function provideGetFullText() {
 224                 return [
 225                         [ NS_MAIN, 'Foo_Bar', '', 'en', 'Foo Bar' ],
 226                         [ NS_USER, 'Hansi_Maier', 'stuff_and_so_on', 'en', 'User:Hansi Maier#stuff and so on' ],
 227
 228                         // No capitalization or normalization is applied while formatting!
 229                         [ NS_USER_TALK, 'hansi__maier', '', 'en', 'User talk:hansi  maier' ],
 230                 ];
 231         }
 232
 233         /**
 234          * @dataProvider provideGetFullText
 235          */
 236         public function testGetFullText( $namespace, $dbkey, $fragment, $lang, $expected ) {
 237                 $codec = $this->makeCodec( $lang );
 238                 $title = new TitleValue( $namespace, $dbkey, $fragment );
 239
 240                 $actual = $codec->getFullText( $title );
 241
 242                 $this->assertEquals( $expected, $actual );
 243         }
 244
 245         public static function provideParseTitle() {
 246                 // TODO: test capitalization and trimming
 247                 // TODO: test unicode normalization
 248
 249                 return [
 250                         [ '  : Hansi_Maier _ ', NS_MAIN, 'en',
 251                                 new TitleValue( NS_MAIN, 'Hansi_Maier', '' ) ],
 252                         [ 'User:::1', NS_MAIN, 'de',
 253                                 new TitleValue( NS_USER, '0:0:0:0:0:0:0:1', '' ) ],
 254                         [ ' lisa Müller', NS_USER, 'de',
 255                                 new TitleValue( NS_USER, 'Lisa_Müller', '' ) ],
 256                         [ 'benutzerin:lisa Müller#stuff', NS_MAIN, 'de',
 257                                 new TitleValue( NS_USER, 'Lisa_Müller', 'stuff' ) ],
 258
 259                         [ ':Category:Quux', NS_MAIN, 'en',
 260                                 new TitleValue( NS_CATEGORY, 'Quux', '' ) ],
 261                         [ 'Category:Quux', NS_MAIN, 'en',
 262                                 new TitleValue( NS_CATEGORY, 'Quux', '' ) ],
 263                         [ 'Category:Quux', NS_CATEGORY, 'en',
 264                                 new TitleValue( NS_CATEGORY, 'Quux', '' ) ],
 265                         [ 'Quux', NS_CATEGORY, 'en',
 266                                 new TitleValue( NS_CATEGORY, 'Quux', '' ) ],
 267                         [ ':Quux', NS_CATEGORY, 'en',
 268                                 new TitleValue( NS_MAIN, 'Quux', '' ) ],
 269
 270                         // getGenderCache() provides a mock that considers first
 271                         // names ending in "a" to be female.
 272
 273                         [ 'a b c', NS_MAIN, 'en',
 274                                 new TitleValue( NS_MAIN, 'A_b_c' ) ],
 275                         [ ' a  b  c ', NS_MAIN, 'en',
 276                                 new TitleValue( NS_MAIN, 'A_b_c' ) ],
 277                         [ ' _ Foo __ Bar_ _', NS_MAIN, 'en',
 278                                 new TitleValue( NS_MAIN, 'Foo_Bar' ) ],
 279
 280                         // NOTE: cases copied from TitleTest::testSecureAndSplit. Keep in sync.
 281                         [ 'Sandbox', NS_MAIN, 'en', ],
 282                         [ 'A "B"', NS_MAIN, 'en', ],
 283                         [ 'A \'B\'', NS_MAIN, 'en', ],
 284                         [ '.com', NS_MAIN, 'en', ],
 285                         [ '~', NS_MAIN, 'en', ],
 286                         [ '"', NS_MAIN, 'en', ],
 287                         [ '\'', NS_MAIN, 'en', ],
 288
 289                         [ 'Talk:Sandbox', NS_MAIN, 'en',
 290                                 new TitleValue( NS_TALK, 'Sandbox' ) ],
 291                         [ 'Talk:Foo:Sandbox', NS_MAIN, 'en',
 292                                 new TitleValue( NS_TALK, 'Foo:Sandbox' ) ],
 293                         [ 'File:Example.svg', NS_MAIN, 'en',
 294                                 new TitleValue( NS_FILE, 'Example.svg' ) ],
 295                         [ 'File_talk:Example.svg', NS_MAIN, 'en',
 296                                 new TitleValue( NS_FILE_TALK, 'Example.svg' ) ],
 297                         [ 'Foo/.../Sandbox', NS_MAIN, 'en',
 298                                 'Foo/.../Sandbox' ],
 299                         [ 'Sandbox/...', NS_MAIN, 'en',
 300                                 'Sandbox/...' ],
 301                         [ 'A~~', NS_MAIN, 'en',
 302                                 'A~~' ],
 303                         // Length is 256 total, but only title part matters
 304                         [ 'Category:' . str_repeat( 'x', 248 ), NS_MAIN, 'en',
 305                                 new TitleValue( NS_CATEGORY,
 306                                         'X' . str_repeat( 'x', 247 ) ) ],
 307                         [ str_repeat( 'x', 252 ), NS_MAIN, 'en',
 308                                 'X' . str_repeat( 'x', 251 ) ]
 309                 ];
 310         }
 311
 312         /**
 313          * @dataProvider provideParseTitle
 314          */
 315         public function testParseTitle( $text, $ns, $lang, $title = null ) {
 316                 if ( $title === null ) {
 317                         $title = str_replace( ' ', '_', trim( $text ) );
 318                 }
 319
 320                 if ( is_string( $title ) ) {
 321                         $title = new TitleValue( NS_MAIN, $title, '' );
 322                 }
 323
 324                 $codec = $this->makeCodec( $lang );
 325                 $actual = $codec->parseTitle( $text, $ns );
 326
 327                 $this->assertEquals( $title, $actual );
 328         }
 329
 330         public static function provideParseTitle_invalid() {
 331                 // TODO: test unicode errors
 332
 333                 return [
 334                         [ '#' ],
 335                         [ '::' ],
 336                         [ '::xx' ],
 337                         [ '::##' ],
 338                         [ ' :: x' ],
 339
 340                         [ 'Talk:File:Foo.jpg' ],
 341                         [ 'Talk:localtestiw:Foo' ],
 342                         [ '::1' ], // only valid in user namespace
 343                         [ 'User::x' ], // leading ":" in a user name is only valid of IPv6 addresses
 344
 345                         // NOTE: cases copied from TitleTest::testSecureAndSplit. Keep in sync.
 346                         [ '' ],
 347                         [ ':' ],
 348                         [ '__  __' ],
 349                         [ '  __  ' ],
 350                         // Bad characters forbidden regardless of wgLegalTitleChars
 351                         [ 'A [ B' ],
 352                         [ 'A ] B' ],
 353                         [ 'A { B' ],
 354                         [ 'A } B' ],
 355                         [ 'A < B' ],
 356                         [ 'A > B' ],
 357                         [ 'A | B' ],
 358                         // URL encoding
 359                         [ 'A%20B' ],
 360                         [ 'A%23B' ],
 361                         [ 'A%2523B' ],
 362                         // XML/HTML character entity references
 363                         // Note: Commented out because they are not marked invalid by the PHP test as
 364                         // Title::newFromText runs Sanitizer::decodeCharReferencesAndNormalize first.
 365                         // [ 'A &eacute; B' ],
 366                         // [ 'A &#233; B' ],
 367                         // [ 'A &#x00E9; B' ],
 368                         // Subject of NS_TALK does not roundtrip to NS_MAIN
 369                         [ 'Talk:File:Example.svg' ],
 370                         // Directory navigation
 371                         [ '.' ],
 372                         [ '..' ],
 373                         [ './Sandbox' ],
 374                         [ '../Sandbox' ],
 375                         [ 'Foo/./Sandbox' ],
 376                         [ 'Foo/../Sandbox' ],
 377                         [ 'Sandbox/.' ],
 378                         [ 'Sandbox/..' ],
 379                         // Tilde
 380                         [ 'A ~~~ Name' ],
 381                         [ 'A ~~~~ Signature' ],
 382                         [ 'A ~~~~~ Timestamp' ],
 383                         [ str_repeat( 'x', 256 ) ],
 384                         // Namespace prefix without actual title
 385                         [ 'Talk:' ],
 386                         [ 'Category: ' ],
 387                         [ 'Category: #bar' ]
 388                 ];
 389         }
 390
 391         /**
 392          * @dataProvider provideParseTitle_invalid
 393          */
 394         public function testParseTitle_invalid( $text ) {
 395                 $this->setExpectedException( MalformedTitleException::class );
 396
 397                 $codec = $this->makeCodec( 'en' );
 398                 $codec->parseTitle( $text, NS_MAIN );
 399         }
 400
 401         public static function provideGetNamespaceName() {
 402                 return [
 403                         [ NS_MAIN, 'Foo', 'en', '' ],
 404                         [ NS_USER, 'Foo', 'en', 'User' ],
 405                         [ NS_USER, 'Hansi Maier', 'de', 'Benutzer' ],
 406
 407                         // getGenderCache() provides a mock that considers first
 408                         // names ending in "a" to be female.
 409                         [ NS_USER, 'Lisa Müller', 'de', 'Benutzerin' ],
 410                 ];
 411         }
 412
 413         /**
 414          * @dataProvider provideGetNamespaceName
 415          */
 416         public function testGetNamespaceName( $namespace, $text, $lang, $expected ) {
 417                 $codec = $this->makeCodec( $lang );
 418                 $name = $codec->getNamespaceName( $namespace, $text );
 419
 420                 $this->assertEquals( $expected, $name );
 421         }
 422 }