includes/externalstore/ExternalStore.php

   1 <?php
   2 /**
   3  * @defgroup ExternalStorage ExternalStorage
   4  */
   5
   6 /**
   7  * Interface for data storage in external repositories.
   8  *
   9  * This program is free software; you can redistribute it and/or modify
  10  * it under the terms of the GNU General Public License as published by
  11  * the Free Software Foundation; either version 2 of the License, or
  12  * (at your option) any later version.
  13  *
  14  * This program is distributed in the hope that it will be useful,
  15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  17  * GNU General Public License for more details.
  18  *
  19  * You should have received a copy of the GNU General Public License along
  20  * with this program; if not, write to the Free Software Foundation, Inc.,
  21  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  22  * http://www.gnu.org/copyleft/gpl.html
  23  *
  24  * @file
  25  */
  26
  27 /**
  28  * Constructor class for key/value blob data kept in external repositories.
  29  *
  30  * Objects in external stores are defined by a special URL. The URL is of
  31  * the form "<store protocol>://<location>/<object name>". The protocol is used
  32  * to determine what ExternalStoreMedium class is used. The location identifies
  33  * particular storage instances or database clusters for store class to use.
  34  *
  35  * When an object is inserted into a store, the calling code uses a partial URL of
  36  * the form "<store protocol>://<location>" and receives the full object URL on success.
  37  * This is useful since object names can be sequential IDs, UUIDs, or hashes.
  38  * Callers are not responsible for unique name generation.
  39  *
  40  * External repositories might be populated by maintenance/async
  41  * scripts, thus partial moving of data may be possible, as well
  42  * as the possibility to have any storage format (i.e. for archives).
  43  *
  44  * @ingroup ExternalStorage
  45  */
  46 class ExternalStore {
  47         /**
  48          * Get an external store object of the given type, with the given parameters
  49          *
  50          * @param string $proto Type of external storage, should be a value in $wgExternalStores
  51          * @param array $params Associative array of ExternalStoreMedium parameters
  52          * @return ExternalStoreMedium|bool The store class or false on error
  53          */
  54         public static function getStoreObject( $proto, array $params = [] ) {
  55                 global $wgExternalStores;
  56
  57                 if ( !$wgExternalStores || !in_array( $proto, $wgExternalStores ) ) {
  58                         return false; // protocol not enabled
  59                 }
  60
  61                 $class = 'ExternalStore' . ucfirst( $proto );
  62
  63                 // Any custom modules should be added to $wgAutoLoadClasses for on-demand loading
  64                 return class_exists( $class ) ? new $class( $params ) : false;
  65         }
  66
  67         /**
  68          * Fetch data from given URL
  69          *
  70          * @param string $url The URL of the text to get
  71          * @param array $params Associative array of ExternalStoreMedium parameters
  72          * @return string|bool The text stored or false on error
  73          * @throws MWException
  74          */
  75         public static function fetchFromURL( $url, array $params = [] ) {
  76                 $parts = explode( '://', $url, 2 );
  77                 if ( count( $parts ) != 2 ) {
  78                         return false; // invalid URL
  79                 }
  80
  81                 list( $proto, $path ) = $parts;
  82                 if ( $path == '' ) { // bad URL
  83                         return false;
  84                 }
  85
  86                 $store = self::getStoreObject( $proto, $params );
  87                 if ( $store === false ) {
  88                         return false;
  89                 }
  90
  91                 return $store->fetchFromURL( $url );
  92         }
  93
  94         /**
  95          * Fetch data from multiple URLs with a minimum of round trips
  96          *
  97          * @param array $urls The URLs of the text to get
  98          * @return array Map from url to its data.  Data is either string when found
  99          *     or false on failure.
 100          */
 101         public static function batchFetchFromURLs( array $urls ) {
 102                 $batches = [];
 103                 foreach ( $urls as $url ) {
 104                         $scheme = parse_url( $url, PHP_URL_SCHEME );
 105                         if ( $scheme ) {
 106                                 $batches[$scheme][] = $url;
 107                         }
 108                 }
 109                 $retval = [];
 110                 foreach ( $batches as $proto => $batchedUrls ) {
 111                         $store = self::getStoreObject( $proto );
 112                         if ( $store === false ) {
 113                                 continue;
 114                         }
 115                         $retval += $store->batchFetchFromURLs( $batchedUrls );
 116                 }
 117                 // invalid, not found, db dead, etc.
 118                 $missing = array_diff( $urls, array_keys( $retval ) );
 119                 if ( $missing ) {
 120                         foreach ( $missing as $url ) {
 121                                 $retval[$url] = false;
 122                         }
 123                 }
 124
 125                 return $retval;
 126         }
 127
 128         /**
 129          * Store a data item to an external store, identified by a partial URL
 130          * The protocol part is used to identify the class, the rest is passed to the
 131          * class itself as a parameter.
 132          *
 133          * @param string $url A partial external store URL ("<store type>://<location>")
 134          * @param string $data
 135          * @param array $params Associative array of ExternalStoreMedium parameters
 136          * @return string|bool The URL of the stored data item, or false on error
 137          * @throws MWException
 138          */
 139         public static function insert( $url, $data, array $params = [] ) {
 140                 $parts = explode( '://', $url, 2 );
 141                 if ( count( $parts ) != 2 ) {
 142                         return false; // invalid URL
 143                 }
 144
 145                 list( $proto, $path ) = $parts;
 146                 if ( $path == '' ) { // bad URL
 147                         return false;
 148                 }
 149
 150                 $store = self::getStoreObject( $proto, $params );
 151                 if ( $store === false ) {
 152                         return false;
 153                 } else {
 154                         return $store->store( $path, $data );
 155                 }
 156         }
 157
 158         /**
 159          * Like insert() above, but does more of the work for us.
 160          * This function does not need a url param, it builds it by
 161          * itself. It also fails-over to the next possible clusters
 162          * provided by $wgDefaultExternalStore.
 163          *
 164          * @param string $data
 165          * @param array $params Associative array of ExternalStoreMedium parameters
 166          * @return string|bool The URL of the stored data item, or false on error
 167          * @throws MWException
 168          */
 169         public static function insertToDefault( $data, array $params = [] ) {
 170                 global $wgDefaultExternalStore;
 171
 172                 return self::insertWithFallback( (array)$wgDefaultExternalStore, $data, $params );
 173         }
 174
 175         /**
 176          * Like insert() above, but does more of the work for us.
 177          * This function does not need a url param, it builds it by
 178          * itself. It also fails-over to the next possible clusters
 179          * as provided in the first parameter.
 180          *
 181          * @param array $tryStores Refer to $wgDefaultExternalStore
 182          * @param string $data
 183          * @param array $params Associative array of ExternalStoreMedium parameters
 184          * @return string|bool The URL of the stored data item, or false on error
 185          * @throws MWException
 186          */
 187         public static function insertWithFallback( array $tryStores, $data, array $params = [] ) {
 188                 $error = false;
 189                 while ( count( $tryStores ) > 0 ) {
 190                         $index = mt_rand( 0, count( $tryStores ) - 1 );
 191                         $storeUrl = $tryStores[$index];
 192                         wfDebug( __METHOD__ . ": trying $storeUrl\n" );
 193                         list( $proto, $path ) = explode( '://', $storeUrl, 2 );
 194                         $store = self::getStoreObject( $proto, $params );
 195                         if ( $store === false ) {
 196                                 throw new MWException( "Invalid external storage protocol - $storeUrl" );
 197                         }
 198                         try {
 199                                 $url = $store->store( $path, $data ); // Try to save the object
 200                         } catch ( Exception $error ) {
 201                                 $url = false;
 202                         }
 203                         if ( strlen( $url ) ) {
 204                                 return $url; // Done!
 205                         } else {
 206                                 unset( $tryStores[$index] ); // Don't try this one again!
 207                                 $tryStores = array_values( $tryStores ); // Must have consecutive keys
 208                                 wfDebugLog( 'ExternalStorage',
 209                                         "Unable to store text to external storage $storeUrl" );
 210                         }
 211                 }
 212                 // All stores failed
 213                 if ( $error ) {
 214                         throw $error; // rethrow the last error
 215                 } else {
 216                         throw new MWException( "Unable to store text to external storage" );
 217                 }
 218         }
 219
 220         /**
 221          * @param string $data
 222          * @param string $wiki
 223          * @return string|bool The URL of the stored data item, or false on error
 224          * @throws MWException
 225          */
 226         public static function insertToForeignDefault( $data, $wiki ) {
 227                 return self::insertToDefault( $data, [ 'wiki' => $wiki ] );
 228         }
 229 }