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 string|bool $pretty If a string, add non-significant whitespace to improve
  99          *   readability, using that string for indentation. If true, use the default indent
 100          *   string (four spaces).
 101          * @param int $escaping Bitfield consisting of _OK class constants
 102          * @return string|bool: String if successful; false upon failure
 103          */
 104         public static function encode( $value, $pretty = false, $escaping = 0 ) {
 105                 if ( !is_string( $pretty ) ) {
 106                         $pretty = $pretty ? '    ' : false;
 107                 }
 108
 109                 if ( defined( 'JSON_UNESCAPED_UNICODE' ) ) {
 110                         return self::encode54( $value, $pretty, $escaping );
 111                 }
 112
 113                 return self::encode53( $value, $pretty, $escaping );
 114         }
 115
 116         /**
 117          * Decodes a JSON string.
 118          *
 119          * @param string $value The JSON string being decoded
 120          * @param bool $assoc When true, returned objects will be converted into associative arrays.
 121          *
 122          * @return mixed The value encoded in JSON in appropriate PHP type.
 123          * `null` is returned if the JSON cannot be decoded or if the encoded data is deeper than
 124          * the recursion limit.
 125          */
 126         public static function decode( $value, $assoc = false ) {
 127                 return json_decode( $value, $assoc );
 128         }
 129
 130         /**
 131          * JSON encoder wrapper for PHP >= 5.4, which supports useful encoding options.
 132          *
 133          * @param mixed $value
 134          * @param string|bool $pretty
 135          * @param int $escaping
 136          * @return string|bool
 137          */
 138         private static function encode54( $value, $pretty, $escaping ) {
 139                 // PHP escapes '/' to prevent breaking out of inline script blocks using '</script>',
 140                 // which is hardly useful when '<' and '>' are escaped (and inadequate), and such
 141                 // escaping negatively impacts the human readability of URLs and similar strings.
 142                 $options = JSON_UNESCAPED_SLASHES;
 143                 $options |= $pretty !== false ? JSON_PRETTY_PRINT : 0;
 144                 $options |= ( $escaping & self::UTF8_OK ) ? JSON_UNESCAPED_UNICODE : 0;
 145                 $options |= ( $escaping & self::XMLMETA_OK ) ? 0 : ( JSON_HEX_TAG | JSON_HEX_AMP );
 146                 $json = json_encode( $value, $options );
 147                 if ( $json === false ) {
 148                         return false;
 149                 }
 150
 151                 if ( $pretty !== false ) {
 152                         // Remove whitespace inside empty arrays/objects; different JSON encoders
 153                         // vary on this, and we want our output to be consistent across implementations.
 154                         $json = preg_replace( self::WS_CLEANUP_REGEX, '', $json );
 155                         if ( $pretty !== '    ' ) {
 156                                 // Change the four-space indent to a tab indent
 157                                 $json = str_replace( "\n    ", "\n\t", $json );
 158                                 while ( strpos( $json, "\t    " ) !== false ) {
 159                                         $json = str_replace( "\t    ", "\t\t", $json );
 160                                 }
 161
 162                                 if ( $pretty !== "\t" ) {
 163                                         // Change the tab indent to the provided indent
 164                                         $json = str_replace( "\t", $pretty, $json );
 165                                 }
 166                         }
 167                 }
 168                 if ( $escaping & self::UTF8_OK ) {
 169                         $json = str_replace( self::$badChars, self::$badCharsEscaped, $json );
 170                 }
 171
 172                 return $json;
 173         }
 174
 175         /**
 176          * JSON encoder wrapper for PHP 5.3, which lacks native support for some encoding options.
 177          * Therefore, the missing options are implemented here purely in PHP code.
 178          *
 179          * @param mixed $value
 180          * @param string|bool $pretty
 181          * @param int $escaping
 182          * @return string|bool
 183          */
 184         private static function encode53( $value, $pretty, $escaping ) {
 185                 $options = ( $escaping & self::XMLMETA_OK ) ? 0 : ( JSON_HEX_TAG | JSON_HEX_AMP );
 186                 $json = json_encode( $value, $options );
 187                 if ( $json === false ) {
 188                         return false;
 189                 }
 190
 191                 // Emulate JSON_UNESCAPED_SLASHES. Because the JSON contains no unescaped slashes
 192                 // (only escaped slashes), a simple string replacement works fine.
 193                 $json = str_replace( '\/', '/', $json );
 194
 195                 if ( $escaping & self::UTF8_OK ) {
 196                         // JSON hex escape sequences follow the format \uDDDD, where DDDD is four hex digits
 197                         // indicating the equivalent UTF-16 code unit's value. To most efficiently unescape
 198                         // them, we exploit the JSON extension's built-in decoder.
 199                         // * We escape the input a second time, so any such sequence becomes \\uDDDD.
 200                         // * To avoid interpreting escape sequences that were in the original input,
 201                         //   each double-escaped backslash (\\\\) is replaced with \\\u005c.
 202                         // * We strip one of the backslashes from each of the escape sequences to unescape.
 203                         // * Then the JSON decoder can perform the actual unescaping.
 204                         $json = str_replace( "\\\\\\\\", "\\\\\\u005c", addcslashes( $json, '\"' ) );
 205                         $json = json_decode( preg_replace( "/\\\\\\\\u(?!00[0-7])/", "\\\\u", "\"$json\"" ) );
 206                         $json = str_replace( self::$badChars, self::$badCharsEscaped, $json );
 207                 }
 208
 209                 if ( $pretty !== false ) {
 210                         return self::prettyPrint( $json, $pretty );
 211                 }
 212
 213                 return $json;
 214         }
 215
 216         /**
 217          * Adds non-significant whitespace to an existing JSON representation of an object.
 218          * Only needed for PHP < 5.4, which lacks the JSON_PRETTY_PRINT option.
 219          *
 220          * @param string $json
 221          * @param string $indentString
 222          * @return string
 223          */
 224         private static function prettyPrint( $json, $indentString ) {
 225                 $buf = '';
 226                 $indent = 0;
 227                 $json = strtr( $json, array( '\\\\' => '\\\\', '\"' => "\x01" ) );
 228                 for ( $i = 0, $n = strlen( $json ); $i < $n; $i += $skip ) {
 229                         $skip = 1;
 230                         switch ( $json[$i] ) {
 231                                 case ':':
 232                                         $buf .= ': ';
 233                                         break;
 234                                 case '[':
 235                                 case '{':
 236                                         ++$indent;
 237                                         // falls through
 238                                 case ',':
 239                                         $buf .= $json[$i] . "\n" . str_repeat( $indentString, $indent );
 240                                         break;
 241                                 case ']':
 242                                 case '}':
 243                                         $buf .= "\n" . str_repeat( $indentString, --$indent ) . $json[$i];
 244                                         break;
 245                                 case '"':
 246                                         $skip = strcspn( $json, '"', $i + 1 ) + 2;
 247                                         $buf .= substr( $json, $i, $skip );
 248                                         break;
 249                                 default:
 250                                         $skip = strcspn( $json, ',]}"', $i + 1 ) + 1;
 251                                         $buf .= substr( $json, $i, $skip );
 252                         }
 253                 }
 254                 $buf = preg_replace( self::WS_CLEANUP_REGEX, '', $buf );
 255
 256                 return str_replace( "\x01", '\"', $buf );
 257         }
 258 }