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