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 namespace MediaWiki\Json;
  24
  25 use MediaWiki\Status\Status;
  26
  27 /**
  28  * JSON formatter wrapper class
  29  */
  30 class FormatJson {
  31         /**
  32          * Skip escaping most characters above U+007F for readability and compactness.
  33          * This encoding option saves 3 to 8 bytes (uncompressed) for each such character;
  34          * however, it could break compatibility with systems that incorrectly handle UTF-8.
  35          *
  36          * @since 1.22
  37          */
  38         public const UTF8_OK = 1;
  39
  40         /**
  41          * Skip escaping the characters '<', '>', and '&', which have special meanings in
  42          * HTML and XML.
  43          *
  44          * @warning Do not use this option for JSON that could end up in inline scripts.
  45          * - HTML 5.2, §4.12.1.3 Restrictions for contents of script elements
  46          * - XML 1.0 (5th Ed.), §2.4 Character Data and Markup
  47          *
  48          * @since 1.22
  49          */
  50         public const XMLMETA_OK = 2;
  51
  52         /**
  53          * Skip escaping as many characters as reasonably possible.
  54          *
  55          * @warning When generating inline script blocks, use FormatJson::UTF8_OK instead.
  56          *
  57          * @since 1.22
  58          */
  59         public const ALL_OK = self::UTF8_OK | self::XMLMETA_OK;
  60
  61         /**
  62          * If set, treat JSON objects '{...}' as associative arrays. Without this option,
  63          * JSON objects will be converted to stdClass.
  64          *
  65          * @since 1.24
  66          */
  67         public const FORCE_ASSOC = 0x100;
  68
  69         /**
  70          * If set, attempt to fix invalid JSON.
  71          *
  72          * @since 1.24
  73          */
  74         public const TRY_FIXING = 0x200;
  75
  76         /**
  77          * If set, strip comments from input before parsing as JSON.
  78          *
  79          * @since 1.25
  80          */
  81         public const STRIP_COMMENTS = 0x400;
  82
  83         /**
  84          * Returns the JSON representation of a value.
  85          *
  86          * @note Empty arrays are encoded as numeric arrays, not as objects, so cast any associative
  87          *       array that might be empty to an object before encoding it.
  88          *
  89          * @note In pre-1.22 versions of MediaWiki, using this function for generating inline script
  90          *       blocks may result in an XSS vulnerability, and quite likely will in XML documents
  91          *       (cf. FormatJson::XMLMETA_OK). Use Xml::encodeJsVar() instead in such cases.
  92          *
  93          * @param mixed $value The value to encode. Can be any type except a resource.
  94          * @param string|bool $pretty If a string, add non-significant whitespace to improve
  95          *   readability, using that string for indentation (must consist only of whitespace
  96          *   characters). If true, use the default indent string (four spaces).
  97          * @param int $escaping Bitfield consisting of _OK class constants
  98          * @return string|false String if successful; false upon failure
  99          */
 100         public static function encode( $value, $pretty = false, $escaping = 0 ) {
 101                 // PHP escapes '/' to prevent breaking out of inline script blocks using '</script>',
 102                 // which is hardly useful when '<' and '>' are escaped (and inadequate), and such
 103                 // escaping negatively impacts the human readability of URLs and similar strings.
 104                 $options = JSON_UNESCAPED_SLASHES;
 105                 if ( $pretty || is_string( $pretty ) ) {
 106                         $options |= JSON_PRETTY_PRINT;
 107                 }
 108                 if ( $escaping & self::UTF8_OK ) {
 109                         $options |= JSON_UNESCAPED_UNICODE;
 110                 }
 111                 if ( !( $escaping & self::XMLMETA_OK ) ) {
 112                         $options |= JSON_HEX_TAG | JSON_HEX_AMP;
 113                 }
 114                 $json = json_encode( $value, $options );
 115
 116                 if ( is_string( $pretty ) && $pretty !== '    ' && $json !== false ) {
 117                         // Change the four-space indent to the provided indent.
 118                         // The regex matches four spaces either at the start of a line or immediately
 119                         // after the previous match. $pretty should contain only whitespace characters,
 120                         // so there should be no need to call StringUtils::escapeRegexReplacement().
 121                         $json = preg_replace( '/ {4}|.*+\n\K {4}/A', $pretty, $json );
 122                 }
 123
 124                 return $json;
 125         }
 126
 127         /**
 128          * Decodes a JSON string. It is recommended to use FormatJson::parse(),
 129          * which returns more comprehensive result in case of an error, and has
 130          * more parsing options.
 131          *
 132          * In PHP versions before 7.1, decoding a JSON string containing an empty key
 133          * without passing $assoc as true results in a return object with a property
 134          * named "_empty_" (because true empty properties were not supported pre-PHP-7.1).
 135          * Instead, consider passing $assoc as true to return an associative array.
 136          *
 137          * But be aware that in all supported PHP versions, decoding an empty JSON object
 138          * with $assoc = true returns an array, not an object, breaking round-trip consistency.
 139          *
 140          * See https://phabricator.wikimedia.org/T206411 for more details on these quirks.
 141          *
 142          * @param string $value The JSON string being decoded
 143          * @param bool $assoc When true, returned objects will be converted into associative arrays.
 144          *
 145          * @return mixed The value encoded in JSON in appropriate PHP type.
 146          * `null` is returned if $value represented `null`, if $value could not be decoded,
 147          * or if the encoded data was deeper than the recursion limit.
 148          * Use FormatJson::parse() to distinguish between types of `null` and to get proper error code.
 149          */
 150         public static function decode( $value, $assoc = false ) {
 151                 return json_decode( $value, $assoc );
 152         }
 153
 154         /**
 155          * Decodes a JSON string.
 156          * Unlike FormatJson::decode(), if $value represents null value, it will be
 157          * properly decoded as valid.
 158          *
 159          * @param string $value The JSON string being decoded
 160          * @param int $options A bit field that allows FORCE_ASSOC, TRY_FIXING,
 161          * STRIP_COMMENTS
 162          * @return Status If valid JSON, the value is available in $result->getValue()
 163          */
 164         public static function parse( $value, $options = 0 ) {
 165                 if ( $options & self::STRIP_COMMENTS ) {
 166                         $value = self::stripComments( $value );
 167                 }
 168                 $assoc = ( $options & self::FORCE_ASSOC ) !== 0;
 169                 $result = json_decode( $value, $assoc );
 170                 $code = json_last_error();
 171
 172                 if ( $code === JSON_ERROR_SYNTAX && ( $options & self::TRY_FIXING ) !== 0 ) {
 173                         // The most common error is the trailing comma in a list or an object.
 174                         // We cannot simply replace /,\s*[}\]]/ because it could be inside a string value.
 175                         // But we could use the fact that JSON does not allow multi-line string values,
 176                         // And remove trailing commas if they are et the end of a line.
 177                         // JSON only allows 4 control characters: [ \t\r\n].  So we must not use '\s' for matching.
 178                         // Regex match   ,]<any non-quote chars>\n   or   ,\n]   with optional spaces/tabs.
 179                         $count = 0;
 180                         $value =
 181                                 preg_replace( '/,([ \t]*[}\]][^"\r\n]*([\r\n]|$)|[ \t]*[\r\n][ \t\r\n]*[}\]])/', '$1',
 182                                         $value, -1, $count );
 183                         if ( $count > 0 ) {
 184                                 $result = json_decode( $value, $assoc );
 185                                 if ( json_last_error() === JSON_ERROR_NONE ) {
 186                                         // Report warning
 187                                         $st = Status::newGood( $result );
 188                                         $st->warning( wfMessage( 'json-warn-trailing-comma' )->numParams( $count ) );
 189                                         return $st;
 190                                 }
 191                         }
 192                 }
 193
 194                 // JSON_ERROR_RECURSION, JSON_ERROR_INF_OR_NAN, JSON_ERROR_UNSUPPORTED_TYPE,
 195                 // are all encode errors that we don't need to care about here.
 196                 switch ( $code ) {
 197                         case JSON_ERROR_NONE:
 198                                 return Status::newGood( $result );
 199                         default:
 200                                 return Status::newFatal( wfMessage( 'json-error-unknown' )->numParams( $code ) );
 201                         case JSON_ERROR_DEPTH:
 202                                 $msg = 'json-error-depth';
 203                                 break;
 204                         case JSON_ERROR_STATE_MISMATCH:
 205                                 $msg = 'json-error-state-mismatch';
 206                                 break;
 207                         case JSON_ERROR_CTRL_CHAR:
 208                                 $msg = 'json-error-ctrl-char';
 209                                 break;
 210                         case JSON_ERROR_SYNTAX:
 211                                 $msg = 'json-error-syntax';
 212                                 break;
 213                         case JSON_ERROR_UTF8:
 214                                 $msg = 'json-error-utf8';
 215                                 break;
 216                         case JSON_ERROR_INVALID_PROPERTY_NAME:
 217                                 $msg = 'json-error-invalid-property-name';
 218                                 break;
 219                         case JSON_ERROR_UTF16:
 220                                 $msg = 'json-error-utf16';
 221                                 break;
 222                 }
 223                 return Status::newFatal( $msg );
 224         }
 225
 226         /**
 227          * Remove multiline and single line comments from an otherwise valid JSON
 228          * input string. This can be used as a preprocessor, to allow JSON
 229          * formatted configuration files to contain comments.
 230          *
 231          * @param string $json
 232          * @return string JSON with comments removed
 233          */
 234         public static function stripComments( $json ) {
 235                 // Ensure we have a string
 236                 $str = (string)$json;
 237                 $buffer = '';
 238                 $maxLen = strlen( $str );
 239                 $mark = 0;
 240
 241                 $inString = false;
 242                 $inComment = false;
 243                 $multiline = false;
 244
 245                 for ( $idx = 0; $idx < $maxLen; $idx++ ) {
 246                         switch ( $str[$idx] ) {
 247                                 case '"':
 248                                         $lookBehind = ( $idx - 1 >= 0 ) ? $str[$idx - 1] : '';
 249                                         if ( !$inComment && $lookBehind !== '\\' ) {
 250                                                 // Either started or ended a string
 251                                                 $inString = !$inString;
 252                                         }
 253                                         break;
 254
 255                                 case '/':
 256                                         $lookAhead = ( $idx + 1 < $maxLen ) ? $str[$idx + 1] : '';
 257                                         $lookBehind = ( $idx - 1 >= 0 ) ? $str[$idx - 1] : '';
 258                                         if ( $inString ) {
 259                                                 break;
 260
 261                                         } elseif ( !$inComment &&
 262                                                 ( $lookAhead === '/' || $lookAhead === '*' )
 263                                         ) {
 264                                                 // Transition into a comment
 265                                                 // Add characters seen to buffer
 266                                                 $buffer .= substr( $str, $mark, $idx - $mark );
 267                                                 // Consume the look ahead character
 268                                                 $idx++;
 269                                                 // Track state
 270                                                 $inComment = true;
 271                                                 $multiline = $lookAhead === '*';
 272
 273                                         } elseif ( $multiline && $lookBehind === '*' ) {
 274                                                 // Found the end of the current comment
 275                                                 $mark = $idx + 1;
 276                                                 $inComment = false;
 277                                                 $multiline = false;
 278                                         }
 279                                         break;
 280
 281                                 case "\n":
 282                                         if ( $inComment && !$multiline ) {
 283                                                 // Found the end of the current comment
 284                                                 $mark = $idx + 1;
 285                                                 $inComment = false;
 286                                         }
 287                                         break;
 288                         }
 289                 }
 290                 if ( $inComment ) {
 291                         // Comment ends with input
 292                         // Technically we should check to ensure that we aren't in
 293                         // a multiline comment that hasn't been properly ended, but this
 294                         // is a strip filter, not a validating parser.
 295                         $mark = $maxLen;
 296                 }
 297                 // Add final chunk to buffer before returning
 298                 return $buffer . substr( $str, $mark, $maxLen - $mark );
 299         }
 300 }
 301 /** @deprecated class alias since 1.43 */
 302 class_alias( FormatJson::class, 'FormatJson' );