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