includes/api/ApiFormatXml.php

   1 <?php
   2 /**
   3  *
   4  *
   5  * Created on Sep 19, 2006
   6  *
   7  * Copyright © 2006 Yuri Astrakhan "<Firstname><Lastname>@gmail.com"
   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  * API XML output formatter
  29  * @ingroup API
  30  */
  31 class ApiFormatXml extends ApiFormatBase {
  32
  33         private $mRootElemName = 'api';
  34         public static $namespace = 'http://www.mediawiki.org/xml/api/';
  35         private $mDoubleQuote = false;
  36         private $mIncludeNamespace = false;
  37         private $mXslt = null;
  38
  39         public function __construct( $main, $format ) {
  40                 parent::__construct( $main, $format );
  41         }
  42
  43         public function getMimeType() {
  44                 return 'text/xml';
  45         }
  46
  47         public function getNeedsRawData() {
  48                 return true;
  49         }
  50
  51         public function setRootElement( $rootElemName ) {
  52                 $this->mRootElemName = $rootElemName;
  53         }
  54
  55         public function execute() {
  56                 $params = $this->extractRequestParams();
  57                 $this->mDoubleQuote = $params['xmldoublequote'];
  58                 $this->mIncludeNamespace = $params['includexmlnamespace'];
  59                 $this->mXslt = $params['xslt'];
  60
  61                 $this->printText( '<?xml version="1.0"?>' );
  62                 if ( !is_null( $this->mXslt ) ) {
  63                         $this->addXslt();
  64                 }
  65                 if ( $this->mIncludeNamespace ) {
  66                         // If the result data already contains an 'xmlns' namespace added
  67                         // for custom XML output types, it will override the one for the
  68                         // generic API results.
  69                         // This allows API output of other XML types like Atom, RSS, RSD.
  70                         $data = $this->getResultData() + array( 'xmlns' => self::$namespace );
  71                 } else {
  72                         $data = $this->getResultData();
  73                 }
  74
  75                 $this->printText(
  76                         self::recXmlPrint( $this->mRootElemName,
  77                                 $data,
  78                                 $this->getIsHtml() ? - 2 : null,
  79                                 $this->mDoubleQuote
  80                         )
  81                 );
  82         }
  83
  84         /**
  85          * This method takes an array and converts it to XML.
  86          *
  87          * There are several noteworthy cases:
  88          *
  89          * If array contains a key '_element', then the code assumes that ALL
  90          * other keys are not important and replaces them with the
  91          * value['_element'].
  92          *
  93          * @par Example:
  94          * @verbatim
  95          * name='root',  value = array( '_element'=>'page', 'x', 'y', 'z')
  96          * @endverbatim
  97          * creates:
  98          * @verbatim
  99          * <root>  <page>x</page>  <page>y</page>  <page>z</page> </root>
 100          * @endverbatim
 101          *
 102          * If any of the array's element key is '*', then the code treats all
 103          * other key->value pairs as attributes, and the value['*'] as the
 104          * element's content.
 105          *
 106          * @par Example:
 107          * @verbatim
 108          * name='root',  value = array( '*'=>'text', 'lang'=>'en', 'id'=>10)
 109          * @endverbatim
 110          * creates:
 111          * @verbatim
 112          * <root lang='en' id='10'>text</root>
 113          * @endverbatim
 114          *
 115          * Finally neither key is found, all keys become element names, and values
 116          * become element content.
 117          *
 118          * @note The method is recursive, so the same rules apply to any
 119          * sub-arrays.
 120          *
 121          * @param $elemName
 122          * @param $elemValue
 123          * @param $indent
 124          * @param $doublequote bool
 125          *
 126          * @return string
 127          */
 128         public static function recXmlPrint( $elemName, $elemValue, $indent, $doublequote = false ) {
 129                 $retval = '';
 130                 if ( !is_null( $indent ) ) {
 131                         $indent += 2;
 132                         $indstr = "\n" . str_repeat( ' ', $indent );
 133                 } else {
 134                         $indstr = '';
 135                 }
 136                 $elemName = str_replace( ' ', '_', $elemName );
 137
 138                 switch ( gettype( $elemValue ) ) {
 139                         case 'array':
 140                                 if ( isset( $elemValue['*'] ) ) {
 141                                         $subElemContent = $elemValue['*'];
 142                                         if ( $doublequote ) {
 143                                                 $subElemContent = Sanitizer::encodeAttribute( $subElemContent );
 144                                         }
 145                                         unset( $elemValue['*'] );
 146
 147                                         // Add xml:space="preserve" to the
 148                                         // element so XML parsers will leave
 149                                         // whitespace in the content alone
 150                                         $elemValue['xml:space'] = 'preserve';
 151                                 } else {
 152                                         $subElemContent = null;
 153                                 }
 154
 155                                 if ( isset( $elemValue['_element'] ) ) {
 156                                         $subElemIndName = $elemValue['_element'];
 157                                         unset( $elemValue['_element'] );
 158                                 } else {
 159                                         $subElemIndName = null;
 160                                 }
 161
 162                                 $indElements = array();
 163                                 $subElements = array();
 164                                 foreach ( $elemValue as $subElemId => & $subElemValue ) {
 165                                         if ( is_string( $subElemValue ) && $doublequote ) {
 166                                                 $subElemValue = Sanitizer::encodeAttribute( $subElemValue );
 167                                         }
 168
 169                                         if ( gettype( $subElemId ) === 'integer' ) {
 170                                                 $indElements[] = $subElemValue;
 171                                                 unset( $elemValue[$subElemId] );
 172                                         } elseif ( is_array( $subElemValue ) ) {
 173                                                 $subElements[$subElemId] = $subElemValue;
 174                                                 unset ( $elemValue[$subElemId] );
 175                                         }
 176                                 }
 177
 178                                 if ( is_null( $subElemIndName ) && count( $indElements ) ) {
 179                                         ApiBase::dieDebug( __METHOD__, "($elemName, ...) has integer keys without _element value. Use ApiResult::setIndexedTagName()." );
 180                                 }
 181
 182                                 if ( count( $subElements ) && count( $indElements ) && !is_null( $subElemContent ) ) {
 183                                         ApiBase::dieDebug( __METHOD__, "($elemName, ...) has content and subelements" );
 184                                 }
 185
 186                                 if ( !is_null( $subElemContent ) ) {
 187                                         $retval .= $indstr . Xml::element( $elemName, $elemValue, $subElemContent );
 188                                 } elseif ( !count( $indElements ) && !count( $subElements ) ) {
 189                                         $retval .= $indstr . Xml::element( $elemName, $elemValue );
 190                                 } else {
 191                                         $retval .= $indstr . Xml::element( $elemName, $elemValue, null );
 192
 193                                         foreach ( $subElements as $subElemId => & $subElemValue ) {
 194                                                 $retval .= self::recXmlPrint( $subElemId, $subElemValue, $indent );
 195                                         }
 196
 197                                         foreach ( $indElements as &$subElemValue ) {
 198                                                 $retval .= self::recXmlPrint( $subElemIndName, $subElemValue, $indent );
 199                                         }
 200
 201                                         $retval .= $indstr . Xml::closeElement( $elemName );
 202                                 }
 203                                 break;
 204                         case 'object':
 205                                 // ignore
 206                                 break;
 207                         default:
 208                                 // to make sure null value doesn't produce unclosed element,
 209                                 // which is what Xml::element( $elemName, null, null ) returns
 210                                 if ( $elemValue === null ) {
 211                                         $retval .= $indstr . Xml::element( $elemName );
 212                                 } else {
 213                                         $retval .= $indstr . Xml::element( $elemName, null, $elemValue );
 214                                 }
 215                                 break;
 216                 }
 217                 return $retval;
 218         }
 219
 220         function addXslt() {
 221                 $nt = Title::newFromText( $this->mXslt );
 222                 if ( is_null( $nt ) || !$nt->exists() ) {
 223                         $this->setWarning( 'Invalid or non-existent stylesheet specified' );
 224                         return;
 225                 }
 226                 if ( $nt->getNamespace() != NS_MEDIAWIKI ) {
 227                         $this->setWarning( 'Stylesheet should be in the MediaWiki namespace.' );
 228                         return;
 229                 }
 230                 if ( substr( $nt->getText(), - 4 ) !== '.xsl' ) {
 231                         $this->setWarning( 'Stylesheet should have .xsl extension.' );
 232                         return;
 233                 }
 234                 $this->printText( '<?xml-stylesheet href="' . htmlspecialchars( $nt->getLocalURL( 'action=raw' ) ) . '" type="text/xsl" ?>' );
 235         }
 236
 237         public function getAllowedParams() {
 238                 return array(
 239                         'xmldoublequote' => false,
 240                         'xslt' => null,
 241                         'includexmlnamespace' => false,
 242                 );
 243         }
 244
 245         public function getParamDescription() {
 246                 return array(
 247                         'xmldoublequote' => 'If specified, double quotes all attributes and content',
 248                         'xslt' => 'If specified, adds <xslt> as stylesheet. This should be a wiki page '
 249                                 . 'in the MediaWiki namespace whose page name ends with ".xsl"',
 250                         'includexmlnamespace' => 'If specified, adds an XML namespace'
 251                 );
 252         }
 253
 254         public function getDescription() {
 255                 return 'Output data in XML format' . parent::getDescription();
 256         }
 257 }