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