includes/json/FormatJson.php

   1 <?php
   2 /**
   3  * Wrapper for json_encode and json_decode.
   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  * JSON formatter wrapper class
  25  */
  26 class FormatJson {
  27         /**
  28          * Skip escaping most characters above U+007F for readability and compactness.
  29          * This encoding option saves 3 to 8 bytes (uncompressed) for each such character;
  30          * however, it could break compatibility with systems that incorrectly handle UTF-8.
  31          *
  32          * @since 1.22
  33          */
  34         const UTF8_OK = 1;
  35
  36         /**
  37          * Skip escaping the characters '<', '>', and '&', which have special meanings in
  38          * HTML and XML.
  39          *
  40          * @warning Do not use this option for JSON that could end up in inline scripts.
  41          * - HTML5, §4.3.1.2 Restrictions for contents of script elements
  42          * - XML 1.0 (5th Ed.), §2.4 Character Data and Markup
  43          *
  44          * @since 1.22
  45          */
  46         const XMLMETA_OK = 2;
  47
  48         /**
  49          * Skip escaping as many characters as reasonably possible.
  50          *
  51          * @warning When generating inline script blocks, use FormatJson::UTF8_OK instead.
  52          *
  53          * @since 1.22
  54          */
  55         const ALL_OK = 3;
  56
  57         /**
  58          * If set, treat json objects '{...}' as associative arrays. Without this option,
  59          * json objects will be converted to stdClass.
  60          * The value is set to 1 to be backward compatible with 'true' that was used before.
  61          *
  62          * @since 1.24
  63          */
  64         const FORCE_ASSOC = 0x100;
  65
  66         /**
  67          * If set, attempts to fix invalid json.
  68          *
  69          * @since 1.24
  70          */
  71         const TRY_FIXING = 0x200;
  72
  73         /**
  74          * Regex that matches whitespace inside empty arrays and objects.
  75          *
  76          * This doesn't affect regular strings inside the JSON because those can't
  77          * have a real line break (\n) in them, at this point they are already escaped
  78          * as the string "\n" which this doesn't match.
  79          *
  80          * @private
  81          */
  82         const WS_CLEANUP_REGEX = '/(?<=[\[{])\n\s*+(?=[\]}])/';
  83
  84         /**
  85          * Characters problematic in JavaScript.
  86          *
  87          * @note These are listed in ECMA-262 (5.1 Ed.), §7.3 Line Terminators along with U+000A (LF)
  88          *       and U+000D (CR). However, PHP already escapes LF and CR according to RFC 4627.
  89          */
  90         private static $badChars = array(
  91                 "\xe2\x80\xa8", // U+2028 LINE SEPARATOR
  92                 "\xe2\x80\xa9", // U+2029 PARAGRAPH SEPARATOR
  93         );
  94
  95         /**
  96          * Escape sequences for characters listed in FormatJson::$badChars.
  97          */
  98         private static $badCharsEscaped = array(
  99                 '\u2028', // U+2028 LINE SEPARATOR
 100                 '\u2029', // U+2029 PARAGRAPH SEPARATOR
 101         );
 102
 103         /**
 104          * Returns the JSON representation of a value.
 105          *
 106          * @note Empty arrays are encoded as numeric arrays, not as objects, so cast any associative
 107          *       array that might be empty to an object before encoding it.
 108          *
 109          * @note In pre-1.22 versions of MediaWiki, using this function for generating inline script
 110          *       blocks may result in an XSS vulnerability, and quite likely will in XML documents
 111          *       (cf. FormatJson::XMLMETA_OK). Use Xml::encodeJsVar() instead in such cases.
 112          *
 113          * @param mixed $value The value to encode. Can be any type except a resource.
 114          * @param string|bool $pretty If a string, add non-significant whitespace to improve
 115          *   readability, using that string for indentation. If true, use the default indent
 116          *   string (four spaces).
 117          * @param int $escaping Bitfield consisting of _OK class constants
 118          * @return string|bool: String if successful; false upon failure
 119          */
 120         public static function encode( $value, $pretty = false, $escaping = 0 ) {
 121                 if ( !is_string( $pretty ) ) {
 122                         $pretty = $pretty ? '    ' : false;
 123                 }
 124
 125                 if ( defined( 'JSON_UNESCAPED_UNICODE' ) ) {
 126                         return self::encode54( $value, $pretty, $escaping );
 127                 }
 128
 129                 return self::encode53( $value, $pretty, $escaping );
 130         }
 131
 132         /**
 133          * Decodes a JSON string. It is recommended to use FormatJson::parse(), which returns more comprehensive
 134          * result in case of an error, and has more parsing options.
 135          *
 136          * @param string $value The JSON string being decoded
 137          * @param bool $assoc When true, returned objects will be converted into associative arrays.
 138          *
 139          * @return mixed The value encoded in JSON in appropriate PHP type.
 140          * `null` is returned if $value represented `null`, if $value could not be decoded,
 141          * or if the encoded data was deeper than the recursion limit.
 142          * Use FormatJson::parse() to distinguish between types of `null` and to get proper error code.
 143          */
 144         public static function decode( $value, $assoc = false ) {
 145                 return json_decode( $value, $assoc );
 146         }
 147
 148         /**
 149          * Decodes a JSON string.
 150          * Unlike FormatJson::decode(), if $value represents null value, it will be properly decoded as valid.
 151          *
 152          * @param string $value The JSON string being decoded
 153          * @param int $options A bit field that allows FORCE_ASSOC, TRY_FIXING
 154          * @return Status If valid JSON, the value is available in $result->getValue()
 155          */
 156         public static function parse( $value, $options = 0 ) {
 157                 $assoc = ( $options & self::FORCE_ASSOC ) !== 0;
 158                 $result = json_decode( $value, $assoc );
 159                 $code = json_last_error();
 160
 161                 if ( $code === JSON_ERROR_SYNTAX && ( $options & self::TRY_FIXING ) !== 0 ) {
 162                         // The most common error is the trailing comma in a list or an object.
 163                         // We cannot simply replace /,\s*[}\]]/ because it could be inside a string value.
 164                         // But we could use the fact that JSON does not allow multi-line string values,
 165                         // And remove trailing commas if they are et the end of a line.
 166                         // JSON only allows 4 control characters: [ \t\r\n].  So we must not use '\s' for matching.
 167                         // Regex match   ,]<any non-quote chars>\n   or   ,\n]   with optional spaces/tabs.
 168                         $count = 0;
 169                         $value =
 170                                 preg_replace( '/,([ \t]*[}\]][^"\r\n]*([\r\n]|$)|[ \t]*[\r\n][ \t\r\n]*[}\]])/', '$1',
 171                                         $value, - 1, $count );
 172                         if ( $count > 0 ) {
 173                                 $result = json_decode( $value, $assoc );
 174                                 if ( JSON_ERROR_NONE === json_last_error() ) {
 175                                         // Report warning
 176                                         $st = Status::newGood( $result );
 177                                         $st->warning( wfMessage( 'json-warn-trailing-comma' )->numParams( $count ) );
 178                                         return $st;
 179                                 }
 180                         }
 181                 }
 182
 183                 switch ( $code ) {
 184                         case JSON_ERROR_NONE:
 185                                 return Status::newGood( $result );
 186                         default:
 187                                 return Status::newFatal( wfMessage( 'json-error-unknown' )->numParams( $code ) );
 188                         case JSON_ERROR_DEPTH:
 189                                 $msg = 'json-error-depth';
 190                                 break;
 191                         case JSON_ERROR_STATE_MISMATCH:
 192                                 $msg = 'json-error-state-mismatch';
 193                                 break;
 194                         case JSON_ERROR_CTRL_CHAR:
 195                                 $msg = 'json-error-ctrl-char';
 196                                 break;
 197                         case JSON_ERROR_SYNTAX:
 198                                 $msg = 'json-error-syntax';
 199                                 break;
 200                         case JSON_ERROR_UTF8:
 201                                 $msg = 'json-error-utf8';
 202                                 break;
 203                         case JSON_ERROR_RECURSION:
 204                                 $msg = 'json-error-recursion';
 205                                 break;
 206                         case JSON_ERROR_INF_OR_NAN:
 207                                 $msg = 'json-error-inf-or-nan';
 208                                 break;
 209                         case JSON_ERROR_UNSUPPORTED_TYPE:
 210                                 $msg = 'json-error-unsupported-type';
 211                                 break;
 212                 }
 213                 return Status::newFatal( $msg );
 214         }
 215
 216         /**
 217          * JSON encoder wrapper for PHP >= 5.4, which supports useful encoding options.
 218          *
 219          * @param mixed $value
 220          * @param string|bool $pretty
 221          * @param int $escaping
 222          * @return string|bool
 223          */
 224         private static function encode54( $value, $pretty, $escaping ) {
 225                 static $bug66021;
 226                 if ( $pretty !== false && $bug66021 === null ) {
 227                         $bug66021 = json_encode( array(), JSON_PRETTY_PRINT ) !== '[]';
 228                 }
 229
 230                 // PHP escapes '/' to prevent breaking out of inline script blocks using '</script>',
 231                 // which is hardly useful when '<' and '>' are escaped (and inadequate), and such
 232                 // escaping negatively impacts the human readability of URLs and similar strings.
 233                 $options = JSON_UNESCAPED_SLASHES;
 234                 $options |= $pretty !== false ? JSON_PRETTY_PRINT : 0;
 235                 $options |= ( $escaping & self::UTF8_OK ) ? JSON_UNESCAPED_UNICODE : 0;
 236                 $options |= ( $escaping & self::XMLMETA_OK ) ? 0 : ( JSON_HEX_TAG | JSON_HEX_AMP );
 237                 $json = json_encode( $value, $options );
 238                 if ( $json === false ) {
 239                         return false;
 240                 }
 241
 242                 if ( $pretty !== false ) {
 243                         // Workaround for <https://bugs.php.net/bug.php?id=66021>
 244                         if ( $bug66021 ) {
 245                                 $json = preg_replace( self::WS_CLEANUP_REGEX, '', $json );
 246                         }
 247                         if ( $pretty !== '    ' ) {
 248                                 // Change the four-space indent to a tab indent
 249                                 $json = str_replace( "\n    ", "\n\t", $json );
 250                                 while ( strpos( $json, "\t    " ) !== false ) {
 251                                         $json = str_replace( "\t    ", "\t\t", $json );
 252                                 }
 253
 254                                 if ( $pretty !== "\t" ) {
 255                                         // Change the tab indent to the provided indent
 256                                         $json = str_replace( "\t", $pretty, $json );
 257                                 }
 258                         }
 259                 }
 260                 if ( $escaping & self::UTF8_OK ) {
 261                         $json = str_replace( self::$badChars, self::$badCharsEscaped, $json );
 262                 }
 263
 264                 return $json;
 265         }
 266
 267         /**
 268          * JSON encoder wrapper for PHP 5.3, which lacks native support for some encoding options.
 269          * Therefore, the missing options are implemented here purely in PHP code.
 270          *
 271          * @param mixed $value
 272          * @param string|bool $pretty
 273          * @param int $escaping
 274          * @return string|bool
 275          */
 276         private static function encode53( $value, $pretty, $escaping ) {
 277                 $options = ( $escaping & self::XMLMETA_OK ) ? 0 : ( JSON_HEX_TAG | JSON_HEX_AMP );
 278                 $json = json_encode( $value, $options );
 279                 if ( $json === false ) {
 280                         return false;
 281                 }
 282
 283                 // Emulate JSON_UNESCAPED_SLASHES. Because the JSON contains no unescaped slashes
 284                 // (only escaped slashes), a simple string replacement works fine.
 285                 $json = str_replace( '\/', '/', $json );
 286
 287                 if ( $escaping & self::UTF8_OK ) {
 288                         // JSON hex escape sequences follow the format \uDDDD, where DDDD is four hex digits
 289                         // indicating the equivalent UTF-16 code unit's value. To most efficiently unescape
 290                         // them, we exploit the JSON extension's built-in decoder.
 291                         // * We escape the input a second time, so any such sequence becomes \\uDDDD.
 292                         // * To avoid interpreting escape sequences that were in the original input,
 293                         //   each double-escaped backslash (\\\\) is replaced with \\\u005c.
 294                         // * We strip one of the backslashes from each of the escape sequences to unescape.
 295                         // * Then the JSON decoder can perform the actual unescaping.
 296                         $json = str_replace( "\\\\\\\\", "\\\\\\u005c", addcslashes( $json, '\"' ) );
 297                         $json = json_decode( preg_replace( "/\\\\\\\\u(?!00[0-7])/", "\\\\u", "\"$json\"" ) );
 298                         $json = str_replace( self::$badChars, self::$badCharsEscaped, $json );
 299                 }
 300
 301                 if ( $pretty !== false ) {
 302                         return self::prettyPrint( $json, $pretty );
 303                 }
 304
 305                 return $json;
 306         }
 307
 308         /**
 309          * Adds non-significant whitespace to an existing JSON representation of an object.
 310          * Only needed for PHP < 5.4, which lacks the JSON_PRETTY_PRINT option.
 311          *
 312          * @param string $json
 313          * @param string $indentString
 314          * @return string
 315          */
 316         private static function prettyPrint( $json, $indentString ) {
 317                 $buf = '';
 318                 $indent = 0;
 319                 $json = strtr( $json, array( '\\\\' => '\\\\', '\"' => "\x01" ) );
 320                 for ( $i = 0, $n = strlen( $json ); $i < $n; $i += $skip ) {
 321                         $skip = 1;
 322                         switch ( $json[$i] ) {
 323                                 case ':':
 324                                         $buf .= ': ';
 325                                         break;
 326                                 case '[':
 327                                 case '{':
 328                                         ++$indent;
 329                                         // falls through
 330                                 case ',':
 331                                         $buf .= $json[$i] . "\n" . str_repeat( $indentString, $indent );
 332                                         break;
 333                                 case ']':
 334                                 case '}':
 335                                         $buf .= "\n" . str_repeat( $indentString, --$indent ) . $json[$i];
 336                                         break;
 337                                 case '"':
 338                                         $skip = strcspn( $json, '"', $i + 1 ) + 2;
 339                                         $buf .= substr( $json, $i, $skip );
 340                                         break;
 341                                 default:
 342                                         $skip = strcspn( $json, ',]}"', $i + 1 ) + 1;
 343                                         $buf .= substr( $json, $i, $skip );
 344                         }
 345                 }
 346                 $buf = preg_replace( self::WS_CLEANUP_REGEX, '', $buf );
 347
 348                 return str_replace( "\x01", '\"', $buf );
 349         }
 350 }