Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
53.75% |
43 / 80 |
|
25.00% |
1 / 4 |
CRAP | |
0.00% |
0 / 1 |
| PopulateSpeechoidLexiconFromWiki | |
56.58% |
43 / 76 |
|
25.00% |
1 / 4 |
48.55 | |
0.00% |
0 / 1 |
| __construct | |
100.00% |
4 / 4 |
|
100.00% |
1 / 1 |
1 | |||
| execute | |
82.22% |
37 / 45 |
|
0.00% |
0 / 1 |
13.95 | |||
| getUser | |
20.00% |
2 / 10 |
|
0.00% |
0 / 1 |
7.61 | |||
| getAllLexiconTitlesForLanguage | |
0.00% |
0 / 17 |
|
0.00% |
0 / 1 |
6 | |||
| 1 | <?php |
| 2 | |
| 3 | namespace MediaWiki\Wikispeech; |
| 4 | |
| 5 | /** |
| 6 | * @file |
| 7 | * @ingroup Extensions |
| 8 | * @license GPL-2.0-or-later |
| 9 | */ |
| 10 | |
| 11 | use Maintenance; |
| 12 | use MediaWiki\MediaWikiServices; |
| 13 | use MediaWiki\Title\Title; |
| 14 | use MediaWiki\User\User; |
| 15 | use MediaWiki\Wikispeech\Lexicon\LexiconSpeechoidStorage; |
| 16 | use MediaWiki\Wikispeech\Lexicon\LexiconWikiStorage; |
| 17 | |
| 18 | /** @var string MediaWiki installation path */ |
| 19 | $IP = getenv( 'MW_INSTALL_PATH' ) ?: __DIR__ . '/../../..'; |
| 20 | require_once "$IP/maintenance/Maintenance.php"; |
| 21 | |
| 22 | /** |
| 23 | * Maintenance script to populate speechoid lexicon |
| 24 | * with entries from wiki. |
| 25 | * |
| 26 | * Be aware that you probably need to execute using mwscript, not php, |
| 27 | * in order to be executed as user www-data, who has access to deleting files. |
| 28 | * |
| 29 | * @since 0.1.13 |
| 30 | */ |
| 31 | class PopulateSpeechoidLexiconFromWiki extends Maintenance { |
| 32 | |
| 33 | /** @var LexiconWikiStorage */ |
| 34 | public $lexiconWikiStorage; |
| 35 | |
| 36 | /** @var LexiconSpeechoidStorage */ |
| 37 | public $speechoidStorage; |
| 38 | |
| 39 | /** @var User */ |
| 40 | public $user; |
| 41 | |
| 42 | /** @var callable|null Used to override titles in tests */ |
| 43 | public $getAllLexiconTitlesForLanguageCallback = null; |
| 44 | |
| 45 | public function __construct() { |
| 46 | parent::__construct(); |
| 47 | $this->requireExtension( 'Wikispeech' ); |
| 48 | $this->addDescription( 'Populate Speechoid storage with entries from local wiki.' ); |
| 49 | $this->addOption( 'user', 'Username to perform edits as', true, true ); |
| 50 | } |
| 51 | |
| 52 | /** |
| 53 | * @return void |
| 54 | */ |
| 55 | public function execute(): void { |
| 56 | $voices = $this->getConfig()->get( 'WikispeechVoices' ); |
| 57 | $languages = array_keys( $voices ); |
| 58 | sort( $languages ); |
| 59 | |
| 60 | $wikiStorage = $this->lexiconWikiStorage ?? WikispeechServices::getLexiconWikiStorage(); |
| 61 | $speechoidStorage = $this->speechoidStorage ?? WikispeechServices::getLexiconSpeechoidStorage(); |
| 62 | |
| 63 | $wikiStorage->setUser( $this->getUser() ); |
| 64 | |
| 65 | foreach ( $languages as $language ) { |
| 66 | $this->output( "Language, $language: " ); |
| 67 | $anyUpdated = false; |
| 68 | |
| 69 | $titles = $this->getAllLexiconTitlesForLanguageCallback |
| 70 | ? call_user_func( $this->getAllLexiconTitlesForLanguageCallback, $language ) |
| 71 | : $this->getAllLexiconTitlesForLanguage( $language ); |
| 72 | |
| 73 | foreach ( $titles as $title ) { |
| 74 | // TODO: This breaks if the lexicon key contains a slash (e.g. "F/V"). |
| 75 | // Consider a more robust way of extracting the key from the title. |
| 76 | $key = explode( '/', $title->getText(), 2 )[1] ?? null; |
| 77 | if ( !$key ) { |
| 78 | continue; |
| 79 | } |
| 80 | |
| 81 | $entry = $wikiStorage->getEntry( $language, $key ); |
| 82 | if ( !$entry ) { |
| 83 | continue; |
| 84 | } |
| 85 | |
| 86 | $updated = false; |
| 87 | $validItems = []; |
| 88 | |
| 89 | foreach ( $entry->getItems() as $item ) { |
| 90 | $identity = $item->getSpeechoidIdentity(); |
| 91 | |
| 92 | if ( $identity === null ) { |
| 93 | $this->output( "Warning: $language/$key has no Speechoid ID — skipping.\n" ); |
| 94 | continue; |
| 95 | } |
| 96 | |
| 97 | $speechoidEntry = $speechoidStorage->getEntry( $language, $key ); |
| 98 | if ( !$speechoidEntry || !$speechoidEntry->findItemBySpeechoidIdentity( $identity ) ) { |
| 99 | $this->output( "Re-creating missing Speechoid entry for $language/$key (ID: $identity)\n" ); |
| 100 | |
| 101 | $item->setSpeechoidIdentity( null ); |
| 102 | |
| 103 | try { |
| 104 | $speechoidStorage->createEntryItem( $language, $key, $item ); |
| 105 | $wikiStorage->replaceEntryItem( $language, $key, $item ); |
| 106 | $this->output( "Re-created Speechoid entry for $language/$key\n" ); |
| 107 | $updated = true; |
| 108 | $anyUpdated = true; |
| 109 | } catch ( \LogicException $e ) { |
| 110 | $this->output( "Failed to re-create entry for $key: " . $e->getMessage() . "\n" ); |
| 111 | } |
| 112 | } |
| 113 | |
| 114 | $validItems[] = $item; |
| 115 | } |
| 116 | |
| 117 | if ( $updated ) { |
| 118 | $newEntry = clone $entry; |
| 119 | $newEntry->setItems( $validItems ); |
| 120 | $wikiStorage->saveLexiconEntryRevision( $language, $key, $newEntry, 'Repaired Speechoid identity' ); |
| 121 | $this->output( "Saved updated wiki entry for $language/$key\n" ); |
| 122 | } |
| 123 | } |
| 124 | |
| 125 | if ( !$anyUpdated ) { |
| 126 | $this->output( "No posts to add\n" ); |
| 127 | } |
| 128 | } |
| 129 | } |
| 130 | |
| 131 | /** |
| 132 | * Helper function to get system user to be able to |
| 133 | * save lexicon entry revision |
| 134 | * |
| 135 | * @return User |
| 136 | */ |
| 137 | protected function getUser(): User { |
| 138 | if ( $this->user !== null ) { |
| 139 | return $this->user; |
| 140 | } |
| 141 | $services = MediaWikiServices::getInstance(); |
| 142 | $username = $this->getOption( 'user' ); |
| 143 | $userIdentity = $services->getUserIdentityLookup()->getUserIdentityByName( $username ); |
| 144 | if ( !$userIdentity ) { |
| 145 | $this->fatalError( |
| 146 | "User '$username' not found or invalid. Please provide an existing username with --user." |
| 147 | ); |
| 148 | } |
| 149 | // @phan-suppress-next-line PhanTypeMismatchArgumentNullable |
| 150 | return $services->getUserFactory()->newFromUserIdentity( $userIdentity ); |
| 151 | } |
| 152 | |
| 153 | /** |
| 154 | * Helper function to get all lexicon titles for a specific language |
| 155 | * |
| 156 | * @param string $language |
| 157 | * @return Title[] |
| 158 | */ |
| 159 | protected function getAllLexiconTitlesForLanguage( string $language ): array { |
| 160 | $titleFactory = MediaWikiServices::getInstance()->getTitleFactory(); |
| 161 | $dbr = MediaWikiServices::getInstance()->getConnectionProvider()->getReplicaDatabase(); |
| 162 | |
| 163 | $prefix = ucfirst( $language ) . '/'; |
| 164 | $like = "$prefix%"; |
| 165 | |
| 166 | $res = $dbr->select( |
| 167 | 'page', |
| 168 | [ 'page_title' ], |
| 169 | [ |
| 170 | 'page_namespace' => NS_PRONUNCIATION_LEXICON, |
| 171 | "page_title LIKE '$like'" |
| 172 | ], |
| 173 | __METHOD__ |
| 174 | ); |
| 175 | |
| 176 | $titles = []; |
| 177 | foreach ( $res as $row ) { |
| 178 | $titles[] = $titleFactory->makeTitle( NS_PRONUNCIATION_LEXICON, $row->page_title ); |
| 179 | } |
| 180 | return $titles; |
| 181 | } |
| 182 | } |
| 183 | |
| 184 | $maintClass = PopulateSpeechoidLexiconFromWiki::class; |
| 185 | require_once RUN_MAINTENANCE_IF_MAIN; |