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 $mIncludeNamespace = false;
  36         private $mXslt = null;
  37
  38         public function getMimeType() {
  39                 return 'text/xml';
  40         }
  41
  42         public function setRootElement( $rootElemName ) {
  43                 $this->mRootElemName = $rootElemName;
  44         }
  45
  46         public function execute() {
  47                 $params = $this->extractRequestParams();
  48                 $this->mIncludeNamespace = $params['includexmlnamespace'];
  49                 $this->mXslt = $params['xslt'];
  50
  51                 $this->printText( '<?xml version="1.0"?>' );
  52                 if ( !is_null( $this->mXslt ) ) {
  53                         $this->addXslt();
  54                 }
  55
  56                 $result = $this->getResult();
  57                 if ( $this->mIncludeNamespace && $result->getResultData( 'xmlns' ) === null ) {
  58                         // If the result data already contains an 'xmlns' namespace added
  59                         // for custom XML output types, it will override the one for the
  60                         // generic API results.
  61                         // This allows API output of other XML types like Atom, RSS, RSD.
  62                         $result->addValue( null, 'xmlns', self::$namespace, ApiResult::NO_SIZE_CHECK );
  63                 }
  64                 $data = $result->getResultData( null, [
  65                         'Custom' => function ( &$data, &$metadata ) {
  66                                 if ( isset( $metadata[ApiResult::META_TYPE] ) ) {
  67                                         // We want to use non-BC for BCassoc to force outputting of _idx.
  68                                         switch ( $metadata[ApiResult::META_TYPE] ) {
  69                                                 case 'BCassoc':
  70                                                         $metadata[ApiResult::META_TYPE] = 'assoc';
  71                                                         break;
  72                                         }
  73                                 }
  74                         },
  75                         'BC' => [ 'nobool', 'no*', 'nosub' ],
  76                         'Types' => [ 'ArmorKVP' => '_name' ],
  77                 ] );
  78
  79                 $this->printText(
  80                         static::recXmlPrint( $this->mRootElemName,
  81                                 $data,
  82                                 $this->getIsHtml() ? -2 : null
  83                         )
  84                 );
  85         }
  86
  87         /**
  88          * This method takes an array and converts it to XML.
  89          *
  90          * @param string|null $name Tag name
  91          * @param mixed $value Tag value (attributes/content/subelements)
  92          * @param int|null $indent Indentation
  93          * @param array $attributes Additional attributes
  94          * @return string
  95          */
  96         public static function recXmlPrint( $name, $value, $indent, $attributes = [] ) {
  97                 $retval = '';
  98                 if ( $indent !== null ) {
  99                         if ( $name !== null ) {
 100                                 $indent += 2;
 101                         }
 102                         $indstr = "\n" . str_repeat( ' ', $indent );
 103                 } else {
 104                         $indstr = '';
 105                 }
 106
 107                 if ( is_object( $value ) ) {
 108                         $value = (array)$value;
 109                 }
 110                 if ( is_array( $value ) ) {
 111                         $contentKey = isset( $value[ApiResult::META_CONTENT] )
 112                                 ? $value[ApiResult::META_CONTENT]
 113                                 : '*';
 114                         $subelementKeys = isset( $value[ApiResult::META_SUBELEMENTS] )
 115                                 ? $value[ApiResult::META_SUBELEMENTS]
 116                                 : [];
 117                         if ( isset( $value[ApiResult::META_BC_SUBELEMENTS] ) ) {
 118                                 $subelementKeys = array_merge(
 119                                         $subelementKeys, $value[ApiResult::META_BC_SUBELEMENTS]
 120                                 );
 121                         }
 122                         $preserveKeys = isset( $value[ApiResult::META_PRESERVE_KEYS] )
 123                                 ? $value[ApiResult::META_PRESERVE_KEYS]
 124                                 : [];
 125                         $indexedTagName = isset( $value[ApiResult::META_INDEXED_TAG_NAME] )
 126                                 ? self::mangleName( $value[ApiResult::META_INDEXED_TAG_NAME], $preserveKeys )
 127                                 : '_v';
 128                         $bcBools = isset( $value[ApiResult::META_BC_BOOLS] )
 129                                 ? $value[ApiResult::META_BC_BOOLS]
 130                                 : [];
 131                         $indexSubelements = isset( $value[ApiResult::META_TYPE] )
 132                                 ? $value[ApiResult::META_TYPE] !== 'array'
 133                                 : false;
 134
 135                         $content = null;
 136                         $subelements = [];
 137                         $indexedSubelements = [];
 138                         foreach ( $value as $k => $v ) {
 139                                 if ( ApiResult::isMetadataKey( $k ) && !in_array( $k, $preserveKeys, true ) ) {
 140                                         continue;
 141                                 }
 142
 143                                 $oldv = $v;
 144                                 if ( is_bool( $v ) && !in_array( $k, $bcBools, true ) ) {
 145                                         $v = $v ? 'true' : 'false';
 146                                 }
 147
 148                                 if ( $name !== null && $k === $contentKey ) {
 149                                         $content = $v;
 150                                 } elseif ( is_int( $k ) ) {
 151                                         $indexedSubelements[$k] = $v;
 152                                 } elseif ( is_array( $v ) || is_object( $v ) ) {
 153                                         $subelements[self::mangleName( $k, $preserveKeys )] = $v;
 154                                 } elseif ( in_array( $k, $subelementKeys, true ) || $name === null ) {
 155                                         $subelements[self::mangleName( $k, $preserveKeys )] = [
 156                                                 'content' => $v,
 157                                                 ApiResult::META_CONTENT => 'content',
 158                                                 ApiResult::META_TYPE => 'assoc',
 159                                         ];
 160                                 } elseif ( is_bool( $oldv ) ) {
 161                                         if ( $oldv ) {
 162                                                 $attributes[self::mangleName( $k, $preserveKeys )] = '';
 163                                         }
 164                                 } elseif ( $v !== null ) {
 165                                         $attributes[self::mangleName( $k, $preserveKeys )] = $v;
 166                                 }
 167                         }
 168
 169                         if ( $content !== null ) {
 170                                 if ( $subelements || $indexedSubelements ) {
 171                                         $subelements[self::mangleName( $contentKey, $preserveKeys )] = [
 172                                                 'content' => $content,
 173                                                 ApiResult::META_CONTENT => 'content',
 174                                                 ApiResult::META_TYPE => 'assoc',
 175                                         ];
 176                                         $content = null;
 177                                 } elseif ( is_scalar( $content ) ) {
 178                                         // Add xml:space="preserve" to the element so XML parsers
 179                                         // will leave whitespace in the content alone
 180                                         $attributes += [ 'xml:space' => 'preserve' ];
 181                                 }
 182                         }
 183
 184                         if ( $content !== null ) {
 185                                 if ( is_scalar( $content ) ) {
 186                                         $retval .= $indstr . Xml::element( $name, $attributes, $content );
 187                                 } else {
 188                                         if ( $name !== null ) {
 189                                                 $retval .= $indstr . Xml::element( $name, $attributes, null );
 190                                         }
 191                                         $retval .= static::recXmlPrint( null, $content, $indent );
 192                                         if ( $name !== null ) {
 193                                                 $retval .= $indstr . Xml::closeElement( $name );
 194                                         }
 195                                 }
 196                         } elseif ( !$indexedSubelements && !$subelements ) {
 197                                 if ( $name !== null ) {
 198                                         $retval .= $indstr . Xml::element( $name, $attributes );
 199                                 }
 200                         } else {
 201                                 if ( $name !== null ) {
 202                                         $retval .= $indstr . Xml::element( $name, $attributes, null );
 203                                 }
 204                                 foreach ( $subelements as $k => $v ) {
 205                                         $retval .= static::recXmlPrint( $k, $v, $indent );
 206                                 }
 207                                 foreach ( $indexedSubelements as $k => $v ) {
 208                                         $retval .= static::recXmlPrint( $indexedTagName, $v, $indent,
 209                                                 $indexSubelements ? [ '_idx' => $k ] : []
 210                                         );
 211                                 }
 212                                 if ( $name !== null ) {
 213                                         $retval .= $indstr . Xml::closeElement( $name );
 214                                 }
 215                         }
 216                 } else {
 217                         // to make sure null value doesn't produce unclosed element,
 218                         // which is what Xml::element( $name, null, null ) returns
 219                         if ( $value === null ) {
 220                                 $retval .= $indstr . Xml::element( $name, $attributes );
 221                         } else {
 222                                 $retval .= $indstr . Xml::element( $name, $attributes, $value );
 223                         }
 224                 }
 225
 226                 return $retval;
 227         }
 228
 229         /**
 230          * Mangle XML-invalid names to be valid in XML
 231          * @param string $name
 232          * @param array $preserveKeys Names to not mangle
 233          * @return string Mangled name
 234          */
 235         private static function mangleName( $name, $preserveKeys = [] ) {
 236                 static $nsc = null, $nc = null;
 237
 238                 if ( in_array( $name, $preserveKeys, true ) ) {
 239                         return $name;
 240                 }
 241
 242                 if ( $name === '' ) {
 243                         return '_';
 244                 }
 245
 246                 if ( $nsc === null ) {
 247                         // Note we omit ':' from $nsc and $nc because it's reserved for XML
 248                         // namespacing, and we omit '_' from $nsc (but not $nc) because we
 249                         // reserve it.
 250                         $nsc = 'A-Za-z\x{C0}-\x{D6}\x{D8}-\x{F6}\x{F8}-\x{2FF}\x{370}-\x{37D}\x{37F}-\x{1FFF}' .
 251                                 '\x{200C}-\x{200D}\x{2070}-\x{218F}\x{2C00}-\x{2FEF}\x{3001}-\x{D7FF}' .
 252                                 '\x{F900}-\x{FDCF}\x{FDF0}-\x{FFFD}\x{10000}-\x{EFFFF}';
 253                         $nc = $nsc . '_\-.0-9\x{B7}\x{300}-\x{36F}\x{203F}-\x{2040}';
 254                 }
 255
 256                 if ( preg_match( "/^[$nsc][$nc]*$/uS", $name ) ) {
 257                         return $name;
 258                 }
 259
 260                 return '_' . preg_replace_callback(
 261                         "/[^$nc]/uS",
 262                         function ( $m ) {
 263                                 return sprintf( '.%X.', UtfNormal\Utils::utf8ToCodepoint( $m[0] ) );
 264                         },
 265                         str_replace( '.', '.2E.', $name )
 266                 );
 267         }
 268
 269         protected function addXslt() {
 270                 $nt = Title::newFromText( $this->mXslt );
 271                 if ( is_null( $nt ) || !$nt->exists() ) {
 272                         $this->addWarning( 'apiwarn-invalidxmlstylesheet' );
 273
 274                         return;
 275                 }
 276                 if ( $nt->getNamespace() != NS_MEDIAWIKI ) {
 277                         $this->addWarning( 'apiwarn-invalidxmlstylesheetns' );
 278
 279                         return;
 280                 }
 281                 if ( substr( $nt->getText(), -4 ) !== '.xsl' ) {
 282                         $this->addWarning( 'apiwarn-invalidxmlstylesheetext' );
 283
 284                         return;
 285                 }
 286                 $this->printText( '<?xml-stylesheet href="' .
 287                         htmlspecialchars( $nt->getLocalURL( 'action=raw' ) ) . '" type="text/xsl" ?>' );
 288         }
 289
 290         public function getAllowedParams() {
 291                 return parent::getAllowedParams() + [
 292                         'xslt' => [
 293                                 ApiBase::PARAM_HELP_MSG => 'apihelp-xml-param-xslt',
 294                         ],
 295                         'includexmlnamespace' => [
 296                                 ApiBase::PARAM_DFLT => false,
 297                                 ApiBase::PARAM_HELP_MSG => 'apihelp-xml-param-includexmlnamespace',
 298                         ],
 299                 ];
 300         }
 301 }