includes/libs/virtualrest/ParsoidVirtualRESTService.php

   1 <?php
   2 /**
   3  * Virtual HTTP service client for Parsoid
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  */
  20
  21 /**
  22  * Virtual REST service for Parsoid
  23  * @since 1.25
  24  */
  25 class ParsoidVirtualRESTService extends VirtualRESTService {
  26         /**
  27          * Example Parsoid v3 requests:
  28          *  GET /local/v3/page/html/$title/{$revision}
  29          *   * $revision is optional
  30          *  POST /local/v3/transform/html/to/wikitext/{$title}{/$revision}
  31          *   * body: array( 'html' => ... )
  32          *   * $title and $revision are optional
  33          *  POST /local/v3/transform/wikitext/to/html/{$title}{/$revision}
  34          *   * body: array( 'wikitext' => ... ) or array( 'wikitext' => ..., 'body_only' => true/false )
  35          *   * $title is optional
  36          *   * $revision is optional
  37          *
  38          * There are also deprecated "v1" requests; see onParsoid1Request
  39          * for details.
  40          * @param array $params Key/value map
  41          *   - url            : Parsoid server URL
  42          *   - domain         : Wiki domain to use
  43          *   - timeout        : Parsoid timeout (optional)
  44          *   - forwardCookies : Cookies to forward to Parsoid, or false. (optional)
  45          *   - HTTPProxy      : Parsoid HTTP proxy (optional)
  46          *   - restbaseCompat : whether to parse URL as if they were meant for RESTBase
  47          *                       boolean (optional)
  48          */
  49         public function __construct( array $params ) {
  50                 // for backwards compatibility:
  51                 if ( isset( $params['URL'] ) ) {
  52                         $params['url'] = $params['URL'];
  53                         unset( $params['URL'] );
  54                 }
  55                 // set up defaults and merge them with the given params
  56                 $mparams = array_merge( [
  57                         'name' => 'parsoid',
  58                         'url' => 'http://localhost:8000/',
  59                         'prefix' => 'localhost',
  60                         'domain' => 'localhost',
  61                         'timeout' => null,
  62                         'forwardCookies' => false,
  63                         'HTTPProxy' => null,
  64                 ], $params );
  65                 // Ensure that the url parameter has a trailing slash.
  66                 if ( substr( $mparams['url'], -1 ) !== '/' ) {
  67                         $mparams['url'] .= '/';
  68                 }
  69                 // Ensure the correct domain format: strip protocol, port,
  70                 // and trailing slash if present.  This lets us use
  71                 // $wgCanonicalServer as a default value, which is very convenient.
  72                 $mparams['domain'] = preg_replace(
  73                         '/^(https?:\/\/)?([^\/:]+?)(:\d+)?\/?$/',
  74                         '$2',
  75                         $mparams['domain']
  76                 );
  77                 parent::__construct( $mparams );
  78         }
  79
  80         public function onRequests( array $reqs, Closure $idGeneratorFunc ) {
  81                 $result = [];
  82                 foreach ( $reqs as $key => $req ) {
  83                         $parts = explode( '/', $req['url'] );
  84
  85                         list(
  86                                 $targetWiki, // 'local'
  87                                 $version, // 'v3' ('v1' for restbase compatibility)
  88                                 $reqType, // 'page' or 'transform'
  89                                 $format, // 'html' or 'wikitext'
  90                                 // $title (optional)
  91                                 // $revision (optional)
  92                         ) = $parts;
  93
  94                         if ( isset( $this->params['restbaseCompat'] ) && $this->params['restbaseCompat'] ) {
  95                                 if ( $version !== 'v1' ) {
  96                                         throw new Exception( "Only RESTBase v1 API is supported." );
  97                                 }
  98                                 # Map RESTBase v1 API to Parsoid v3 API (pretty easy)
  99                                 $req['url'] = preg_replace( '#^local/v1/#', 'local/v3/', $req['url'] );
 100                         } elseif ( $version !== 'v3' ) {
 101                                 $result[$key] = $this->onParsoid1Request( $req, $idGeneratorFunc );
 102                                 continue;
 103                         }
 104                         if ( $targetWiki !== 'local' ) {
 105                                 throw new Exception( "Only 'local' target wiki is currently supported" );
 106                         }
 107                         if ( $reqType !== 'page' && $reqType !== 'transform' ) {
 108                                 throw new Exception( "Request action must be either 'page' or 'transform'" );
 109                         }
 110                         if ( $format !== 'html' && $format !== 'wikitext' ) {
 111                                 throw new Exception( "Request format must be either 'html' or 'wt'" );
 112                         }
 113                         // replace /local/ with the current domain
 114                         $req['url'] = preg_replace( '#^local/#', $this->params['domain'] . '/', $req['url'] );
 115                         // and prefix it with the service URL
 116                         $req['url'] = $this->params['url'] . $req['url'];
 117                         // set the appropriate proxy, timeout and headers
 118                         if ( $this->params['HTTPProxy'] ) {
 119                                 $req['proxy'] = $this->params['HTTPProxy'];
 120                         }
 121                         if ( $this->params['timeout'] != null ) {
 122                                 $req['reqTimeout'] = $this->params['timeout'];
 123                         }
 124                         if ( $this->params['forwardCookies'] ) {
 125                                 $req['headers']['Cookie'] = $this->params['forwardCookies'];
 126                         }
 127                         $result[$key] = $req;
 128                 }
 129                 return $result;
 130         }
 131
 132         /**
 133          * Remap a Parsoid v1 request to a Parsoid v3 request.
 134          *
 135          * Example Parsoid v1 requests:
 136          *  GET /local/v1/page/$title/html/$oldid
 137          *   * $oldid is optional
 138          *  POST /local/v1/transform/html/to/wikitext/$title/$oldid
 139          *   * body: array( 'html' => ... )
 140          *   * $title and $oldid are optional
 141          *  POST /local/v1/transform/wikitext/to/html/$title
 142          *   * body: array( 'wikitext' => ... ) or array( 'wikitext' => ..., 'body' => true/false )
 143          *   * $title is optional
 144          *
 145          * NOTE: the POST APIs aren't "real" Parsoid v1 APIs, they are just what
 146          * Visual Editor "pretends" the V1 API is like.  A previous version of
 147          * ParsoidVirtualRESTService translated these to the "real" Parsoid v1
 148          * API.  We now translate these to the "real" Parsoid v3 API.
 149          * @param array $req
 150          * @param Closure $idGeneratorFunc
 151          * @return array
 152          * @throws Exception
 153          */
 154         public function onParsoid1Request( array $req, Closure $idGeneratorFunc ) {
 155                 $parts = explode( '/', $req['url'] );
 156                 list(
 157                         $targetWiki, // 'local'
 158                         $version, // 'v1'
 159                         $reqType // 'page' or 'transform'
 160                 ) = $parts;
 161                 if ( $targetWiki !== 'local' ) {
 162                         throw new Exception( "Only 'local' target wiki is currently supported" );
 163                 } elseif ( $version !== 'v1' ) {
 164                         throw new Exception( "Only v1 and v3 are supported." );
 165                 } elseif ( $reqType !== 'page' && $reqType !== 'transform' ) {
 166                         throw new Exception( "Request type must be either 'page' or 'transform'" );
 167                 }
 168                 $req['url'] = $this->params['url'] . $this->params['domain'] . '/v3/';
 169                 if ( $reqType === 'page' ) {
 170                         $title = $parts[3];
 171                         if ( $parts[4] !== 'html' ) {
 172                                 throw new Exception( "Only 'html' output format is currently supported" );
 173                         }
 174                         $req['url'] .= 'page/html/' . $title;
 175                         if ( isset( $parts[5] ) ) {
 176                                 $req['url'] .= '/' . $parts[5];
 177                         } elseif ( isset( $req['query']['oldid'] ) && $req['query']['oldid'] ) {
 178                                 $req['url'] .= '/' . $req['query']['oldid'];
 179                                 unset( $req['query']['oldid'] );
 180                         }
 181                 } elseif ( $reqType === 'transform' ) {
 182                         $req['url'] .= 'transform/'. $parts[3] . '/to/' . $parts[5];
 183                         // the title
 184                         if ( isset( $parts[6] ) ) {
 185                                 $req['url'] .= '/' . $parts[6];
 186                         }
 187                         // revision id
 188                         if ( isset( $parts[7] ) ) {
 189                                 $req['url'] .= '/' . $parts[7];
 190                         } elseif ( isset( $req['body']['oldid'] ) && $req['body']['oldid'] ) {
 191                                 $req['url'] .= '/' . $req['body']['oldid'];
 192                                 unset( $req['body']['oldid'] );
 193                         }
 194                         if ( $parts[4] !== 'to' ) {
 195                                 throw new Exception( "Part index 4 is not 'to'" );
 196                         }
 197                         if ( $parts[3] === 'html' && $parts[5] === 'wikitext' ) {
 198                                 if ( !isset( $req['body']['html'] ) ) {
 199                                         throw new Exception( "You must set an 'html' body key for this request" );
 200                                 }
 201                         } elseif ( $parts[3] == 'wikitext' && $parts[5] == 'html' ) {
 202                                 if ( !isset( $req['body']['wikitext'] ) ) {
 203                                         throw new Exception( "You must set a 'wikitext' body key for this request" );
 204                                 }
 205                                 if ( isset( $req['body']['body'] ) ) {
 206                                         $req['body']['body_only'] = $req['body']['body'];
 207                                         unset( $req['body']['body'] );
 208                                 }
 209                         } else {
 210                                 throw new Exception( "Transformation unsupported" );
 211                         }
 212                 }
 213                 // set the appropriate proxy, timeout and headers
 214                 if ( $this->params['HTTPProxy'] ) {
 215                         $req['proxy'] = $this->params['HTTPProxy'];
 216                 }
 217                 if ( $this->params['timeout'] != null ) {
 218                         $req['reqTimeout'] = $this->params['timeout'];
 219                 }
 220                 if ( $this->params['forwardCookies'] ) {
 221                         $req['headers']['Cookie'] = $this->params['forwardCookies'];
 222                 }
 223
 224                 return $req;
 225         }
 226
 227 }