includes/api/ApiFormatXml.php

   1 <?php
   2 /**
   3  * Copyright © 2006 Yuri Astrakhan "<Firstname><Lastname>@gmail.com"
   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  * @file
  21  */
  22
  23 /**
  24  * API XML output formatter
  25  * @ingroup API
  26  */
  27 class ApiFormatXml extends ApiFormatBase {
  28
  29         private $mRootElemName = 'api';
  30         public static $namespace = 'http://www.mediawiki.org/xml/api/';
  31         private $mIncludeNamespace = false;
  32         private $mXslt = null;
  33
  34         public function getMimeType() {
  35                 return 'text/xml';
  36         }
  37
  38         public function setRootElement( $rootElemName ) {
  39                 $this->mRootElemName = $rootElemName;
  40         }
  41
  42         public function execute() {
  43                 $params = $this->extractRequestParams();
  44                 $this->mIncludeNamespace = $params['includexmlnamespace'];
  45                 $this->mXslt = $params['xslt'];
  46
  47                 $this->printText( '<?xml version="1.0"?>' );
  48                 if ( !is_null( $this->mXslt ) ) {
  49                         $this->addXslt();
  50                 }
  51
  52                 $result = $this->getResult();
  53                 if ( $this->mIncludeNamespace && $result->getResultData( 'xmlns' ) === null ) {
  54                         // If the result data already contains an 'xmlns' namespace added
  55                         // for custom XML output types, it will override the one for the
  56                         // generic API results.
  57                         // This allows API output of other XML types like Atom, RSS, RSD.
  58                         $result->addValue( null, 'xmlns', self::$namespace, ApiResult::NO_SIZE_CHECK );
  59                 }
  60                 $data = $result->getResultData( null, [
  61                         'Custom' => function ( &$data, &$metadata ) {
  62                                 if ( isset( $metadata[ApiResult::META_TYPE] ) ) {
  63                                         // We want to use non-BC for BCassoc to force outputting of _idx.
  64                                         switch ( $metadata[ApiResult::META_TYPE] ) {
  65                                                 case 'BCassoc':
  66                                                         $metadata[ApiResult::META_TYPE] = 'assoc';
  67                                                         break;
  68                                         }
  69                                 }
  70                         },
  71                         'BC' => [ 'nobool', 'no*', 'nosub' ],
  72                         'Types' => [ 'ArmorKVP' => '_name' ],
  73                 ] );
  74
  75                 $this->printText(
  76                         static::recXmlPrint( $this->mRootElemName,
  77                                 $data,
  78                                 $this->getIsHtml() ? -2 : null
  79                         )
  80                 );
  81         }
  82
  83         /**
  84          * This method takes an array and converts it to XML.
  85          *
  86          * @param string|null $name Tag name
  87          * @param mixed $value Tag value (attributes/content/subelements)
  88          * @param int|null $indent Indentation
  89          * @param array $attributes Additional attributes
  90          * @return string
  91          */
  92         public static function recXmlPrint( $name, $value, $indent, $attributes = [] ) {
  93                 $retval = '';
  94                 if ( $indent !== null ) {
  95                         if ( $name !== null ) {
  96                                 $indent += 2;
  97                         }
  98                         $indstr = "\n" . str_repeat( ' ', $indent );
  99                 } else {
 100                         $indstr = '';
 101                 }
 102
 103                 if ( is_object( $value ) ) {
 104                         $value = (array)$value;
 105                 }
 106                 if ( is_array( $value ) ) {
 107                         $contentKey = isset( $value[ApiResult::META_CONTENT] )
 108                                 ? $value[ApiResult::META_CONTENT]
 109                                 : '*';
 110                         $subelementKeys = isset( $value[ApiResult::META_SUBELEMENTS] )
 111                                 ? $value[ApiResult::META_SUBELEMENTS]
 112                                 : [];
 113                         if ( isset( $value[ApiResult::META_BC_SUBELEMENTS] ) ) {
 114                                 $subelementKeys = array_merge(
 115                                         $subelementKeys, $value[ApiResult::META_BC_SUBELEMENTS]
 116                                 );
 117                         }
 118                         $preserveKeys = isset( $value[ApiResult::META_PRESERVE_KEYS] )
 119                                 ? $value[ApiResult::META_PRESERVE_KEYS]
 120                                 : [];
 121                         $indexedTagName = isset( $value[ApiResult::META_INDEXED_TAG_NAME] )
 122                                 ? self::mangleName( $value[ApiResult::META_INDEXED_TAG_NAME], $preserveKeys )
 123                                 : '_v';
 124                         $bcBools = isset( $value[ApiResult::META_BC_BOOLS] )
 125                                 ? $value[ApiResult::META_BC_BOOLS]
 126                                 : [];
 127                         $indexSubelements = isset( $value[ApiResult::META_TYPE] )
 128                                 ? $value[ApiResult::META_TYPE] !== 'array'
 129                                 : false;
 130
 131                         $content = null;
 132                         $subelements = [];
 133                         $indexedSubelements = [];
 134                         foreach ( $value as $k => $v ) {
 135                                 if ( ApiResult::isMetadataKey( $k ) && !in_array( $k, $preserveKeys, true ) ) {
 136                                         continue;
 137                                 }
 138
 139                                 $oldv = $v;
 140                                 if ( is_bool( $v ) && !in_array( $k, $bcBools, true ) ) {
 141                                         $v = $v ? 'true' : 'false';
 142                                 }
 143
 144                                 if ( $name !== null && $k === $contentKey ) {
 145                                         $content = $v;
 146                                 } elseif ( is_int( $k ) ) {
 147                                         $indexedSubelements[$k] = $v;
 148                                 } elseif ( is_array( $v ) || is_object( $v ) ) {
 149                                         $subelements[self::mangleName( $k, $preserveKeys )] = $v;
 150                                 } elseif ( in_array( $k, $subelementKeys, true ) || $name === null ) {
 151                                         $subelements[self::mangleName( $k, $preserveKeys )] = [
 152                                                 'content' => $v,
 153                                                 ApiResult::META_CONTENT => 'content',
 154                                                 ApiResult::META_TYPE => 'assoc',
 155                                         ];
 156                                 } elseif ( is_bool( $oldv ) ) {
 157                                         if ( $oldv ) {
 158                                                 $attributes[self::mangleName( $k, $preserveKeys )] = '';
 159                                         }
 160                                 } elseif ( $v !== null ) {
 161                                         $attributes[self::mangleName( $k, $preserveKeys )] = $v;
 162                                 }
 163                         }
 164
 165                         if ( $content !== null ) {
 166                                 if ( $subelements || $indexedSubelements ) {
 167                                         $subelements[self::mangleName( $contentKey, $preserveKeys )] = [
 168                                                 'content' => $content,
 169                                                 ApiResult::META_CONTENT => 'content',
 170                                                 ApiResult::META_TYPE => 'assoc',
 171                                         ];
 172                                         $content = null;
 173                                 } elseif ( is_scalar( $content ) ) {
 174                                         // Add xml:space="preserve" to the element so XML parsers
 175                                         // will leave whitespace in the content alone
 176                                         $attributes += [ 'xml:space' => 'preserve' ];
 177                                 }
 178                         }
 179
 180                         if ( $content !== null ) {
 181                                 if ( is_scalar( $content ) ) {
 182                                         $retval .= $indstr . Xml::element( $name, $attributes, $content );
 183                                 } else {
 184                                         if ( $name !== null ) {
 185                                                 $retval .= $indstr . Xml::element( $name, $attributes, null );
 186                                         }
 187                                         $retval .= static::recXmlPrint( null, $content, $indent );
 188                                         if ( $name !== null ) {
 189                                                 $retval .= $indstr . Xml::closeElement( $name );
 190                                         }
 191                                 }
 192                         } elseif ( !$indexedSubelements && !$subelements ) {
 193                                 if ( $name !== null ) {
 194                                         $retval .= $indstr . Xml::element( $name, $attributes );
 195                                 }
 196                         } else {
 197                                 if ( $name !== null ) {
 198                                         $retval .= $indstr . Xml::element( $name, $attributes, null );
 199                                 }
 200                                 foreach ( $subelements as $k => $v ) {
 201                                         $retval .= static::recXmlPrint( $k, $v, $indent );
 202                                 }
 203                                 foreach ( $indexedSubelements as $k => $v ) {
 204                                         $retval .= static::recXmlPrint( $indexedTagName, $v, $indent,
 205                                                 $indexSubelements ? [ '_idx' => $k ] : []
 206                                         );
 207                                 }
 208                                 if ( $name !== null ) {
 209                                         $retval .= $indstr . Xml::closeElement( $name );
 210                                 }
 211                         }
 212                 } else {
 213                         // to make sure null value doesn't produce unclosed element,
 214                         // which is what Xml::element( $name, null, null ) returns
 215                         if ( $value === null ) {
 216                                 $retval .= $indstr . Xml::element( $name, $attributes );
 217                         } else {
 218                                 $retval .= $indstr . Xml::element( $name, $attributes, $value );
 219                         }
 220                 }
 221
 222                 return $retval;
 223         }
 224
 225         /**
 226          * Mangle XML-invalid names to be valid in XML
 227          * @param string $name
 228          * @param array $preserveKeys Names to not mangle
 229          * @return string Mangled name
 230          */
 231         private static function mangleName( $name, $preserveKeys = [] ) {
 232                 static $nsc = null, $nc = null;
 233
 234                 if ( in_array( $name, $preserveKeys, true ) ) {
 235                         return $name;
 236                 }
 237
 238                 if ( $name === '' ) {
 239                         return '_';
 240                 }
 241
 242                 if ( $nsc === null ) {
 243                         // Note we omit ':' from $nsc and $nc because it's reserved for XML
 244                         // namespacing, and we omit '_' from $nsc (but not $nc) because we
 245                         // reserve it.
 246                         $nsc = 'A-Za-z\x{C0}-\x{D6}\x{D8}-\x{F6}\x{F8}-\x{2FF}\x{370}-\x{37D}\x{37F}-\x{1FFF}' .
 247                                 '\x{200C}-\x{200D}\x{2070}-\x{218F}\x{2C00}-\x{2FEF}\x{3001}-\x{D7FF}' .
 248                                 '\x{F900}-\x{FDCF}\x{FDF0}-\x{FFFD}\x{10000}-\x{EFFFF}';
 249                         $nc = $nsc . '_\-.0-9\x{B7}\x{300}-\x{36F}\x{203F}-\x{2040}';
 250                 }
 251
 252                 if ( preg_match( "/^[$nsc][$nc]*$/uS", $name ) ) {
 253                         return $name;
 254                 }
 255
 256                 return '_' . preg_replace_callback(
 257                         "/[^$nc]/uS",
 258                         function ( $m ) {
 259                                 return sprintf( '.%X.', UtfNormal\Utils::utf8ToCodepoint( $m[0] ) );
 260                         },
 261                         str_replace( '.', '.2E.', $name )
 262                 );
 263         }
 264
 265         protected function addXslt() {
 266                 $nt = Title::newFromText( $this->mXslt );
 267                 if ( is_null( $nt ) || !$nt->exists() ) {
 268                         $this->addWarning( 'apiwarn-invalidxmlstylesheet' );
 269
 270                         return;
 271                 }
 272                 if ( $nt->getNamespace() != NS_MEDIAWIKI ) {
 273                         $this->addWarning( 'apiwarn-invalidxmlstylesheetns' );
 274
 275                         return;
 276                 }
 277                 if ( substr( $nt->getText(), -4 ) !== '.xsl' ) {
 278                         $this->addWarning( 'apiwarn-invalidxmlstylesheetext' );
 279
 280                         return;
 281                 }
 282                 $this->printText( '<?xml-stylesheet href="' .
 283                         htmlspecialchars( $nt->getLocalURL( 'action=raw' ) ) . '" type="text/xsl" ?>' );
 284         }
 285
 286         public function getAllowedParams() {
 287                 return parent::getAllowedParams() + [
 288                         'xslt' => [
 289                                 ApiBase::PARAM_HELP_MSG => 'apihelp-xml-param-xslt',
 290                         ],
 291                         'includexmlnamespace' => [
 292                                 ApiBase::PARAM_DFLT => false,
 293                                 ApiBase::PARAM_HELP_MSG => 'apihelp-xml-param-includexmlnamespace',
 294                         ],
 295                 ];
 296         }
 297 }