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 = 0x1;
  65
  66         /**
  67          * If set, attempts to fix invalid json.
  68          *
  69          * @since 1.24
  70          */
  71         const TRY_FIXING = 0x2;
  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.
 134          *
 135          * @param string $value The JSON string being decoded
 136          * @param bool $assoc When true, returned objects will be converted into associative arrays.
 137          *
 138          * @return mixed The value encoded in JSON in appropriate PHP type.
 139          * `null` is returned if the JSON cannot be decoded or if the encoded data is deeper than
 140          * the recursion limit.
 141          */
 142         public static function decode( $value, $assoc = false ) {
 143                 return json_decode( $value, $assoc );
 144         }
 145
 146         /**
 147          * Decodes a JSON string.
 148          *
 149          * @param string $value The JSON string being decoded
 150          * @param int $options A bit field that allows FORCE_ASSOC, TRY_FIXING, WRAP_RESULT
 151          * For backward compatibility, FORCE_ASSOC is set to 1 to match the legacy 'true'
 152          * @return Status If good, the value is available in $result->getValue()
 153          */
 154         public static function parse( $value, $options = 0 ) {
 155                 $assoc = ( $options & self::FORCE_ASSOC ) !== 0;
 156                 $result = json_decode( $value, $assoc );
 157                 $code = json_last_error();
 158
 159                 if ( $code === JSON_ERROR_SYNTAX && ( $options & self::TRY_FIXING ) !== 0 ) {
 160                         // The most common error is the trailing comma in a list or an object.
 161                         // We cannot simply replace /,\s*[}\]]/ because it could be inside a string value.
 162                         // But we could use the fact that JSON does not allow multi-line string values,
 163                         // And remove trailing commas if they are et the end of a line.
 164                         // JSON only allows 4 control characters: [ \t\r\n].  So we must not use '\s' for matching.
 165                         // Regex match   ,]<any non-quote chars>\n   or   ,\n]   with optional spaces/tabs.
 166                         $count = 0;
 167                         $value =
 168                                 preg_replace( '/,([ \t]*[}\]][^"\r\n]*([\r\n]|$)|[ \t]*[\r\n][ \t\r\n]*[}\]])/', '$1',
 169                                         $value, - 1, $count );
 170                         if ( $count > 0 ) {
 171                                 $result = json_decode( $value, $assoc );
 172                                 if ( JSON_ERROR_NONE === json_last_error() ) {
 173                                         // Report warning
 174                                         $st = Status::newGood( $result );
 175                                         $st->warning( wfMessage( 'json-warn-trailing-comma' )->numParams( $count ) );
 176                                         return $st;
 177                                 }
 178                         }
 179                 }
 180
 181                 switch ( $code ) {
 182                         case JSON_ERROR_NONE:
 183                                 return Status::newGood( $result );
 184                         default:
 185                                 return Status::newFatal( wfMessage( 'json-error-unknown' )->numParams( $code ) );
 186                         case JSON_ERROR_DEPTH:
 187                                 $msg = 'json-error-depth';
 188                                 break;
 189                         case JSON_ERROR_STATE_MISMATCH:
 190                                 $msg = 'json-error-state-mismatch';
 191                                 break;
 192                         case JSON_ERROR_CTRL_CHAR:
 193                                 $msg = 'json-error-ctrl-char';
 194                                 break;
 195                         case JSON_ERROR_SYNTAX:
 196                                 $msg = 'json-error-syntax';
 197                                 break;
 198                         case JSON_ERROR_UTF8:
 199                                 $msg = 'json-error-utf8';
 200                                 break;
 201                         case JSON_ERROR_RECURSION:
 202                                 $msg = 'json-error-recursion';
 203                                 break;
 204                         case JSON_ERROR_INF_OR_NAN:
 205                                 $msg = 'json-error-inf-or-nan';
 206                                 break;
 207                         case JSON_ERROR_UNSUPPORTED_TYPE:
 208                                 $msg = 'json-error-unsupported-type';
 209                                 break;
 210                 }
 211                 return Status::newFatal( $msg );
 212         }
 213
 214         /**
 215          * JSON encoder wrapper for PHP >= 5.4, which supports useful encoding options.
 216          *
 217          * @param mixed $value
 218          * @param string|bool $pretty
 219          * @param int $escaping
 220          * @return string|bool
 221          */
 222         private static function encode54( $value, $pretty, $escaping ) {
 223                 static $bug66021;
 224                 if ( $pretty !== false && $bug66021 === null ) {
 225                         $bug66021 = json_encode( array(), JSON_PRETTY_PRINT ) !== '[]';
 226                 }
 227
 228                 // PHP escapes '/' to prevent breaking out of inline script blocks using '</script>',
 229                 // which is hardly useful when '<' and '>' are escaped (and inadequate), and such
 230                 // escaping negatively impacts the human readability of URLs and similar strings.
 231                 $options = JSON_UNESCAPED_SLASHES;
 232                 $options |= $pretty !== false ? JSON_PRETTY_PRINT : 0;
 233                 $options |= ( $escaping & self::UTF8_OK ) ? JSON_UNESCAPED_UNICODE : 0;
 234                 $options |= ( $escaping & self::XMLMETA_OK ) ? 0 : ( JSON_HEX_TAG | JSON_HEX_AMP );
 235                 $json = json_encode( $value, $options );
 236                 if ( $json === false ) {
 237                         return false;
 238                 }
 239
 240                 if ( $pretty !== false ) {
 241                         // Workaround for <https://bugs.php.net/bug.php?id=66021>
 242                         if ( $bug66021 ) {
 243                                 $json = preg_replace( self::WS_CLEANUP_REGEX, '', $json );
 244                         }
 245                         if ( $pretty !== '    ' ) {
 246                                 // Change the four-space indent to a tab indent
 247                                 $json = str_replace( "\n    ", "\n\t", $json );
 248                                 while ( strpos( $json, "\t    " ) !== false ) {
 249                                         $json = str_replace( "\t    ", "\t\t", $json );
 250                                 }
 251
 252                                 if ( $pretty !== "\t" ) {
 253                                         // Change the tab indent to the provided indent
 254                                         $json = str_replace( "\t", $pretty, $json );
 255                                 }
 256                         }
 257                 }
 258                 if ( $escaping & self::UTF8_OK ) {
 259                         $json = str_replace( self::$badChars, self::$badCharsEscaped, $json );
 260                 }
 261
 262                 return $json;
 263         }
 264
 265         /**
 266          * JSON encoder wrapper for PHP 5.3, which lacks native support for some encoding options.
 267          * Therefore, the missing options are implemented here purely in PHP code.
 268          *
 269          * @param mixed $value
 270          * @param string|bool $pretty
 271          * @param int $escaping
 272          * @return string|bool
 273          */
 274         private static function encode53( $value, $pretty, $escaping ) {
 275                 $options = ( $escaping & self::XMLMETA_OK ) ? 0 : ( JSON_HEX_TAG | JSON_HEX_AMP );
 276                 $json = json_encode( $value, $options );
 277                 if ( $json === false ) {
 278                         return false;
 279                 }
 280
 281                 // Emulate JSON_UNESCAPED_SLASHES. Because the JSON contains no unescaped slashes
 282                 // (only escaped slashes), a simple string replacement works fine.
 283                 $json = str_replace( '\/', '/', $json );
 284
 285                 if ( $escaping & self::UTF8_OK ) {
 286                         // JSON hex escape sequences follow the format \uDDDD, where DDDD is four hex digits
 287                         // indicating the equivalent UTF-16 code unit's value. To most efficiently unescape
 288                         // them, we exploit the JSON extension's built-in decoder.
 289                         // * We escape the input a second time, so any such sequence becomes \\uDDDD.
 290                         // * To avoid interpreting escape sequences that were in the original input,
 291                         //   each double-escaped backslash (\\\\) is replaced with \\\u005c.
 292                         // * We strip one of the backslashes from each of the escape sequences to unescape.
 293                         // * Then the JSON decoder can perform the actual unescaping.
 294                         $json = str_replace( "\\\\\\\\", "\\\\\\u005c", addcslashes( $json, '\"' ) );
 295                         $json = json_decode( preg_replace( "/\\\\\\\\u(?!00[0-7])/", "\\\\u", "\"$json\"" ) );
 296                         $json = str_replace( self::$badChars, self::$badCharsEscaped, $json );
 297                 }
 298
 299                 if ( $pretty !== false ) {
 300                         return self::prettyPrint( $json, $pretty );
 301                 }
 302
 303                 return $json;
 304         }
 305
 306         /**
 307          * Adds non-significant whitespace to an existing JSON representation of an object.
 308          * Only needed for PHP < 5.4, which lacks the JSON_PRETTY_PRINT option.
 309          *
 310          * @param string $json
 311          * @param string $indentString
 312          * @return string
 313          */
 314         private static function prettyPrint( $json, $indentString ) {
 315                 $buf = '';
 316                 $indent = 0;
 317                 $json = strtr( $json, array( '\\\\' => '\\\\', '\"' => "\x01" ) );
 318                 for ( $i = 0, $n = strlen( $json ); $i < $n; $i += $skip ) {
 319                         $skip = 1;
 320                         switch ( $json[$i] ) {
 321                                 case ':':
 322                                         $buf .= ': ';
 323                                         break;
 324                                 case '[':
 325                                 case '{':
 326                                         ++$indent;
 327                                         // falls through
 328                                 case ',':
 329                                         $buf .= $json[$i] . "\n" . str_repeat( $indentString, $indent );
 330                                         break;
 331                                 case ']':
 332                                 case '}':
 333                                         $buf .= "\n" . str_repeat( $indentString, --$indent ) . $json[$i];
 334                                         break;
 335                                 case '"':
 336                                         $skip = strcspn( $json, '"', $i + 1 ) + 2;
 337                                         $buf .= substr( $json, $i, $skip );
 338                                         break;
 339                                 default:
 340                                         $skip = strcspn( $json, ',]}"', $i + 1 ) + 1;
 341                                         $buf .= substr( $json, $i, $skip );
 342                         }
 343                 }
 344                 $buf = preg_replace( self::WS_CLEANUP_REGEX, '', $buf );
 345
 346                 return str_replace( "\x01", '\"', $buf );
 347         }
 348 }