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          * - HTML 5.2, §4.12.1.3 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 = self::UTF8_OK | self::XMLMETA_OK;
  56
  57         /**
  58          * If set, treat JSON objects '{...}' as associative arrays. Without this option,
  59          * JSON objects will be converted to stdClass.
  60          *
  61          * @since 1.24
  62          */
  63         const FORCE_ASSOC = 0x100;
  64
  65         /**
  66          * If set, attempt to fix invalid JSON.
  67          *
  68          * @since 1.24
  69          */
  70         const TRY_FIXING = 0x200;
  71
  72         /**
  73          * If set, strip comments from input before parsing as JSON.
  74          *
  75          * @since 1.25
  76          */
  77         const STRIP_COMMENTS = 0x400;
  78
  79         /**
  80          * Characters problematic in JavaScript.
  81          *
  82          * @note These are listed in ECMA-262 (5.1 Ed.), §7.3 Line Terminators along with U+000A (LF)
  83          *       and U+000D (CR). However, PHP already escapes LF and CR according to RFC 4627.
  84          */
  85         private static $badChars = [
  86                 "\u{2028}", // U+2028 LINE SEPARATOR
  87                 "\u{2029}", // U+2029 PARAGRAPH SEPARATOR
  88         ];
  89
  90         /**
  91          * Escape sequences for characters listed in FormatJson::$badChars.
  92          */
  93         private static $badCharsEscaped = [
  94                 '\u2028', // U+2028 LINE SEPARATOR
  95                 '\u2029', // U+2029 PARAGRAPH SEPARATOR
  96         ];
  97
  98         /**
  99          * Returns the JSON representation of a value.
 100          *
 101          * @note Empty arrays are encoded as numeric arrays, not as objects, so cast any associative
 102          *       array that might be empty to an object before encoding it.
 103          *
 104          * @note In pre-1.22 versions of MediaWiki, using this function for generating inline script
 105          *       blocks may result in an XSS vulnerability, and quite likely will in XML documents
 106          *       (cf. FormatJson::XMLMETA_OK). Use Xml::encodeJsVar() instead in such cases.
 107          *
 108          * @param mixed $value The value to encode. Can be any type except a resource.
 109          * @param string|bool $pretty If a string, add non-significant whitespace to improve
 110          *   readability, using that string for indentation. If true, use the default indent
 111          *   string (four spaces).
 112          * @param int $escaping Bitfield consisting of _OK class constants
 113          * @return string|false String if successful; false upon failure
 114          */
 115         public static function encode( $value, $pretty = false, $escaping = 0 ) {
 116                 if ( !is_string( $pretty ) ) {
 117                         $pretty = $pretty ? '    ' : false;
 118                 }
 119
 120                 // PHP escapes '/' to prevent breaking out of inline script blocks using '</script>',
 121                 // which is hardly useful when '<' and '>' are escaped (and inadequate), and such
 122                 // escaping negatively impacts the human readability of URLs and similar strings.
 123                 $options = JSON_UNESCAPED_SLASHES;
 124                 $options |= $pretty !== false ? JSON_PRETTY_PRINT : 0;
 125                 $options |= ( $escaping & self::UTF8_OK ) ? JSON_UNESCAPED_UNICODE : 0;
 126                 $options |= ( $escaping & self::XMLMETA_OK ) ? 0 : ( JSON_HEX_TAG | JSON_HEX_AMP );
 127                 $json = json_encode( $value, $options );
 128                 if ( $json === false ) {
 129                         return false;
 130                 }
 131
 132                 if ( $pretty !== false && $pretty !== '    ' ) {
 133                         // Change the four-space indent to a tab indent
 134                         $json = str_replace( "\n    ", "\n\t", $json );
 135                         while ( strpos( $json, "\t    " ) !== false ) {
 136                                 $json = str_replace( "\t    ", "\t\t", $json );
 137                         }
 138
 139                         if ( $pretty !== "\t" ) {
 140                                 // Change the tab indent to the provided indent
 141                                 $json = str_replace( "\t", $pretty, $json );
 142                         }
 143                 }
 144                 if ( $escaping & self::UTF8_OK ) {
 145                         $json = str_replace( self::$badChars, self::$badCharsEscaped, $json );
 146                 }
 147
 148                 return $json;
 149         }
 150
 151         /**
 152          * Decodes a JSON string. It is recommended to use FormatJson::parse(),
 153          * which returns more comprehensive result in case of an error, and has
 154          * more parsing options.
 155          *
 156          * In PHP versions before 7.1, decoding a JSON string containing an empty key
 157          * without passing $assoc as true results in a return object with a property
 158          * named "_empty_" (because true empty properties were not supported pre-PHP-7.1).
 159          * Instead, consider passing $assoc as true to return an associative array.
 160          *
 161          * But be aware that in all supported PHP versions, decoding an empty JSON object
 162          * with $assoc = true returns an array, not an object, breaking round-trip consistency.
 163          *
 164          * See https://phabricator.wikimedia.org/T206411 for more details on these quirks.
 165          *
 166          * @param string $value The JSON string being decoded
 167          * @param bool $assoc When true, returned objects will be converted into associative arrays.
 168          *
 169          * @return mixed The value encoded in JSON in appropriate PHP type.
 170          * `null` is returned if $value represented `null`, if $value could not be decoded,
 171          * or if the encoded data was deeper than the recursion limit.
 172          * Use FormatJson::parse() to distinguish between types of `null` and to get proper error code.
 173          */
 174         public static function decode( $value, $assoc = false ) {
 175                 return json_decode( $value, $assoc );
 176         }
 177
 178         /**
 179          * Decodes a JSON string.
 180          * Unlike FormatJson::decode(), if $value represents null value, it will be
 181          * properly decoded as valid.
 182          *
 183          * @param string $value The JSON string being decoded
 184          * @param int $options A bit field that allows FORCE_ASSOC, TRY_FIXING,
 185          * STRIP_COMMENTS
 186          * @return Status If valid JSON, the value is available in $result->getValue()
 187          */
 188         public static function parse( $value, $options = 0 ) {
 189                 if ( $options & self::STRIP_COMMENTS ) {
 190                         $value = self::stripComments( $value );
 191                 }
 192                 $assoc = ( $options & self::FORCE_ASSOC ) !== 0;
 193                 $result = json_decode( $value, $assoc );
 194                 $code = json_last_error();
 195
 196                 if ( $code === JSON_ERROR_SYNTAX && ( $options & self::TRY_FIXING ) !== 0 ) {
 197                         // The most common error is the trailing comma in a list or an object.
 198                         // We cannot simply replace /,\s*[}\]]/ because it could be inside a string value.
 199                         // But we could use the fact that JSON does not allow multi-line string values,
 200                         // And remove trailing commas if they are et the end of a line.
 201                         // JSON only allows 4 control characters: [ \t\r\n].  So we must not use '\s' for matching.
 202                         // Regex match   ,]<any non-quote chars>\n   or   ,\n]   with optional spaces/tabs.
 203                         $count = 0;
 204                         $value =
 205                                 preg_replace( '/,([ \t]*[}\]][^"\r\n]*([\r\n]|$)|[ \t]*[\r\n][ \t\r\n]*[}\]])/', '$1',
 206                                         $value, -1, $count );
 207                         if ( $count > 0 ) {
 208                                 $result = json_decode( $value, $assoc );
 209                                 if ( JSON_ERROR_NONE === json_last_error() ) {
 210                                         // Report warning
 211                                         $st = Status::newGood( $result );
 212                                         $st->warning( wfMessage( 'json-warn-trailing-comma' )->numParams( $count ) );
 213                                         return $st;
 214                                 }
 215                         }
 216                 }
 217
 218                 switch ( $code ) {
 219                         case JSON_ERROR_NONE:
 220                                 return Status::newGood( $result );
 221                         default:
 222                                 return Status::newFatal( wfMessage( 'json-error-unknown' )->numParams( $code ) );
 223                         case JSON_ERROR_DEPTH:
 224                                 $msg = 'json-error-depth';
 225                                 break;
 226                         case JSON_ERROR_STATE_MISMATCH:
 227                                 $msg = 'json-error-state-mismatch';
 228                                 break;
 229                         case JSON_ERROR_CTRL_CHAR:
 230                                 $msg = 'json-error-ctrl-char';
 231                                 break;
 232                         case JSON_ERROR_SYNTAX:
 233                                 $msg = 'json-error-syntax';
 234                                 break;
 235                         case JSON_ERROR_UTF8:
 236                                 $msg = 'json-error-utf8';
 237                                 break;
 238                         case JSON_ERROR_RECURSION:
 239                                 $msg = 'json-error-recursion';
 240                                 break;
 241                         case JSON_ERROR_INF_OR_NAN:
 242                                 $msg = 'json-error-inf-or-nan';
 243                                 break;
 244                         case JSON_ERROR_UNSUPPORTED_TYPE:
 245                                 $msg = 'json-error-unsupported-type';
 246                                 break;
 247                 }
 248                 return Status::newFatal( $msg );
 249         }
 250
 251         /**
 252          * Remove multiline and single line comments from an otherwise valid JSON
 253          * input string. This can be used as a preprocessor, to allow JSON
 254          * formatted configuration files to contain comments.
 255          *
 256          * @param string $json
 257          * @return string JSON with comments removed
 258          */
 259         public static function stripComments( $json ) {
 260                 // Ensure we have a string
 261                 $str = (string)$json;
 262                 $buffer = '';
 263                 $maxLen = strlen( $str );
 264                 $mark = 0;
 265
 266                 $inString = false;
 267                 $inComment = false;
 268                 $multiline = false;
 269
 270                 for ( $idx = 0; $idx < $maxLen; $idx++ ) {
 271                         switch ( $str[$idx] ) {
 272                                 case '"':
 273                                         $lookBehind = ( $idx - 1 >= 0 ) ? $str[$idx - 1] : '';
 274                                         if ( !$inComment && $lookBehind !== '\\' ) {
 275                                                 // Either started or ended a string
 276                                                 $inString = !$inString;
 277                                         }
 278                                         break;
 279
 280                                 case '/':
 281                                         $lookAhead = ( $idx + 1 < $maxLen ) ? $str[$idx + 1] : '';
 282                                         $lookBehind = ( $idx - 1 >= 0 ) ? $str[$idx - 1] : '';
 283                                         if ( $inString ) {
 284                                                 break;
 285
 286                                         } elseif ( !$inComment &&
 287                                                 ( $lookAhead === '/' || $lookAhead === '*' )
 288                                         ) {
 289                                                 // Transition into a comment
 290                                                 // Add characters seen to buffer
 291                                                 $buffer .= substr( $str, $mark, $idx - $mark );
 292                                                 // Consume the look ahead character
 293                                                 $idx++;
 294                                                 // Track state
 295                                                 $inComment = true;
 296                                                 $multiline = $lookAhead === '*';
 297
 298                                         } elseif ( $multiline && $lookBehind === '*' ) {
 299                                                 // Found the end of the current comment
 300                                                 $mark = $idx + 1;
 301                                                 $inComment = false;
 302                                                 $multiline = false;
 303                                         }
 304                                         break;
 305
 306                                 case "\n":
 307                                         if ( $inComment && !$multiline ) {
 308                                                 // Found the end of the current comment
 309                                                 $mark = $idx + 1;
 310                                                 $inComment = false;
 311                                         }
 312                                         break;
 313                         }
 314                 }
 315                 if ( $inComment ) {
 316                         // Comment ends with input
 317                         // Technically we should check to ensure that we aren't in
 318                         // a multiline comment that hasn't been properly ended, but this
 319                         // is a strip filter, not a validating parser.
 320                         $mark = $maxLen;
 321                 }
 322                 // Add final chunk to buffer before returning
 323                 return $buffer . substr( $str, $mark, $maxLen - $mark );
 324         }
 325 }