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          * Regex that matches whitespace inside empty arrays and objects.
  59          *
  60          * This doesn't affect regular strings inside the JSON because those can't
  61          * have a real line break (\n) in them, at this point they are already escaped
  62          * as the string "\n" which this doesn't match.
  63          *
  64          * @private
  65          */
  66         const WS_CLEANUP_REGEX = '/(?<=[\[{])\n\s*+(?=[\]}])/';
  67
  68         /**
  69          * Characters problematic in JavaScript.
  70          *
  71          * @note These are listed in ECMA-262 (5.1 Ed.), §7.3 Line Terminators along with U+000A (LF)
  72          *       and U+000D (CR). However, PHP already escapes LF and CR according to RFC 4627.
  73          */
  74         private static $badChars = array(
  75                 "\xe2\x80\xa8", // U+2028 LINE SEPARATOR
  76                 "\xe2\x80\xa9", // U+2029 PARAGRAPH SEPARATOR
  77         );
  78
  79         /**
  80          * Escape sequences for characters listed in FormatJson::$badChars.
  81          */
  82         private static $badCharsEscaped = array(
  83                 '\u2028', // U+2028 LINE SEPARATOR
  84                 '\u2029', // U+2029 PARAGRAPH SEPARATOR
  85         );
  86
  87         /**
  88          * Returns the JSON representation of a value.
  89          *
  90          * @note Empty arrays are encoded as numeric arrays, not as objects, so cast any associative
  91          *       array that might be empty to an object before encoding it.
  92          *
  93          * @note In pre-1.22 versions of MediaWiki, using this function for generating inline script
  94          *       blocks may result in an XSS vulnerability, and quite likely will in XML documents
  95          *       (cf. FormatJson::XMLMETA_OK). Use Xml::encodeJsVar() instead in such cases.
  96          *
  97          * @param mixed $value The value to encode. Can be any type except a resource.
  98          * @param bool $pretty If true, add non-significant whitespace to improve readability.
  99          * @param int $escaping Bitfield consisting of _OK class constants
 100          * @return string|bool: String if successful; false upon failure
 101          */
 102         public static function encode( $value, $pretty = false, $escaping = 0 ) {
 103                 if ( defined( 'JSON_UNESCAPED_UNICODE' ) ) {
 104                         return self::encode54( $value, $pretty, $escaping );
 105                 }
 106
 107                 return self::encode53( $value, $pretty, $escaping );
 108         }
 109
 110         /**
 111          * Decodes a JSON string.
 112          *
 113          * @param string $value The JSON string being decoded
 114          * @param bool $assoc When true, returned objects will be converted into associative arrays.
 115          *
 116          * @return mixed The value encoded in JSON in appropriate PHP type.
 117          * `null` is returned if the JSON cannot be decoded or if the encoded data is deeper than
 118          * the recursion limit.
 119          */
 120         public static function decode( $value, $assoc = false ) {
 121                 return json_decode( $value, $assoc );
 122         }
 123
 124         /**
 125          * JSON encoder wrapper for PHP >= 5.4, which supports useful encoding options.
 126          *
 127          * @param mixed $value
 128          * @param bool $pretty
 129          * @param int $escaping
 130          * @return string|bool
 131          */
 132         private static function encode54( $value, $pretty, $escaping ) {
 133                 // PHP escapes '/' to prevent breaking out of inline script blocks using '</script>',
 134                 // which is hardly useful when '<' and '>' are escaped (and inadequate), and such
 135                 // escaping negatively impacts the human readability of URLs and similar strings.
 136                 $options = JSON_UNESCAPED_SLASHES;
 137                 $options |= $pretty ? JSON_PRETTY_PRINT : 0;
 138                 $options |= ( $escaping & self::UTF8_OK ) ? JSON_UNESCAPED_UNICODE : 0;
 139                 $options |= ( $escaping & self::XMLMETA_OK ) ? 0 : ( JSON_HEX_TAG | JSON_HEX_AMP );
 140                 $json = json_encode( $value, $options );
 141                 if ( $json === false ) {
 142                         return false;
 143                 }
 144
 145                 if ( $pretty ) {
 146                         // Remove whitespace inside empty arrays/objects; different JSON encoders
 147                         // vary on this, and we want our output to be consistent across implementations.
 148                         $json = preg_replace( self::WS_CLEANUP_REGEX, '', $json );
 149                 }
 150                 if ( $escaping & self::UTF8_OK ) {
 151                         $json = str_replace( self::$badChars, self::$badCharsEscaped, $json );
 152                 }
 153
 154                 return $json;
 155         }
 156
 157         /**
 158          * JSON encoder wrapper for PHP 5.3, which lacks native support for some encoding options.
 159          * Therefore, the missing options are implemented here purely in PHP code.
 160          *
 161          * @param mixed $value
 162          * @param bool $pretty
 163          * @param int $escaping
 164          * @return string|bool
 165          */
 166         private static function encode53( $value, $pretty, $escaping ) {
 167                 $options = ( $escaping & self::XMLMETA_OK ) ? 0 : ( JSON_HEX_TAG | JSON_HEX_AMP );
 168                 $json = json_encode( $value, $options );
 169                 if ( $json === false ) {
 170                         return false;
 171                 }
 172
 173                 // Emulate JSON_UNESCAPED_SLASHES. Because the JSON contains no unescaped slashes
 174                 // (only escaped slashes), a simple string replacement works fine.
 175                 $json = str_replace( '\/', '/', $json );
 176
 177                 if ( $escaping & self::UTF8_OK ) {
 178                         // JSON hex escape sequences follow the format \uDDDD, where DDDD is four hex digits
 179                         // indicating the equivalent UTF-16 code unit's value. To most efficiently unescape
 180                         // them, we exploit the JSON extension's built-in decoder.
 181                         // * We escape the input a second time, so any such sequence becomes \\uDDDD.
 182                         // * To avoid interpreting escape sequences that were in the original input,
 183                         //   each double-escaped backslash (\\\\) is replaced with \\\u005c.
 184                         // * We strip one of the backslashes from each of the escape sequences to unescape.
 185                         // * Then the JSON decoder can perform the actual unescaping.
 186                         $json = str_replace( "\\\\\\\\", "\\\\\\u005c", addcslashes( $json, '\"' ) );
 187                         $json = json_decode( preg_replace( "/\\\\\\\\u(?!00[0-7])/", "\\\\u", "\"$json\"" ) );
 188                         $json = str_replace( self::$badChars, self::$badCharsEscaped, $json );
 189                 }
 190
 191                 if ( $pretty ) {
 192                         return self::prettyPrint( $json );
 193                 }
 194
 195                 return $json;
 196         }
 197
 198         /**
 199          * Adds non-significant whitespace to an existing JSON representation of an object.
 200          * Only needed for PHP < 5.4, which lacks the JSON_PRETTY_PRINT option.
 201          *
 202          * @param string $json
 203          * @return string
 204          */
 205         private static function prettyPrint( $json ) {
 206                 $buf = '';
 207                 $indent = 0;
 208                 $json = strtr( $json, array( '\\\\' => '\\\\', '\"' => "\x01" ) );
 209                 for ( $i = 0, $n = strlen( $json ); $i < $n; $i += $skip ) {
 210                         $skip = 1;
 211                         switch ( $json[$i] ) {
 212                                 case ':':
 213                                         $buf .= ': ';
 214                                         break;
 215                                 case '[':
 216                                 case '{':
 217                                         ++$indent;
 218                                         // falls through
 219                                 case ',':
 220                                         $buf .= $json[$i] . "\n" . str_repeat( '    ', $indent );
 221                                         break;
 222                                 case ']':
 223                                 case '}':
 224                                         $buf .= "\n" . str_repeat( '    ', --$indent ) . $json[$i];
 225                                         break;
 226                                 case '"':
 227                                         $skip = strcspn( $json, '"', $i + 1 ) + 2;
 228                                         $buf .= substr( $json, $i, $skip );
 229                                         break;
 230                                 default:
 231                                         $skip = strcspn( $json, ',]}"', $i + 1 ) + 1;
 232                                         $buf .= substr( $json, $i, $skip );
 233                         }
 234                 }
 235                 $buf = preg_replace( self::WS_CLEANUP_REGEX, '', $buf );
 236
 237                 return str_replace( "\x01", '\"', $buf );
 238         }
 239 }