MediaWiki  1.28.1
FormatJson.php
Go to the documentation of this file.
1 <?php
26 class FormatJson {
34  const UTF8_OK = 1;
35 
46  const XMLMETA_OK = 2;
47 
55  const ALL_OK = 3;
56 
64  const FORCE_ASSOC = 0x100;
65 
71  const TRY_FIXING = 0x200;
72 
78  const STRIP_COMMENTS = 0x400;
79 
89  const WS_CLEANUP_REGEX = '/(?<=[\[{])\n\s*+(?=[\]}])/';
90 
97  private static $badChars = [
98  "\xe2\x80\xa8", // U+2028 LINE SEPARATOR
99  "\xe2\x80\xa9", // U+2029 PARAGRAPH SEPARATOR
100  ];
101 
105  private static $badCharsEscaped = [
106  '\u2028', // U+2028 LINE SEPARATOR
107  '\u2029', // U+2029 PARAGRAPH SEPARATOR
108  ];
109 
127  public static function encode( $value, $pretty = false, $escaping = 0 ) {
128  if ( !is_string( $pretty ) ) {
129  $pretty = $pretty ? ' ' : false;
130  }
131 
132  static $bug66021;
133  if ( $pretty !== false && $bug66021 === null ) {
134  $bug66021 = json_encode( [], JSON_PRETTY_PRINT ) !== '[]';
135  }
136 
137  // PHP escapes '/' to prevent breaking out of inline script blocks using '</script>',
138  // which is hardly useful when '<' and '>' are escaped (and inadequate), and such
139  // escaping negatively impacts the human readability of URLs and similar strings.
140  $options = JSON_UNESCAPED_SLASHES;
141  $options |= $pretty !== false ? JSON_PRETTY_PRINT : 0;
142  $options |= ( $escaping & self::UTF8_OK ) ? JSON_UNESCAPED_UNICODE : 0;
143  $options |= ( $escaping & self::XMLMETA_OK ) ? 0 : ( JSON_HEX_TAG | JSON_HEX_AMP );
144  $json = json_encode( $value, $options );
145  if ( $json === false ) {
146  return false;
147  }
148 
149  if ( $pretty !== false ) {
150  // Workaround for <https://bugs.php.net/bug.php?id=66021>
151  if ( $bug66021 ) {
152  $json = preg_replace( self::WS_CLEANUP_REGEX, '', $json );
153  }
154  if ( $pretty !== ' ' ) {
155  // Change the four-space indent to a tab indent
156  $json = str_replace( "\n ", "\n\t", $json );
157  while ( strpos( $json, "\t " ) !== false ) {
158  $json = str_replace( "\t ", "\t\t", $json );
159  }
160 
161  if ( $pretty !== "\t" ) {
162  // Change the tab indent to the provided indent
163  $json = str_replace( "\t", $pretty, $json );
164  }
165  }
166  }
167  if ( $escaping & self::UTF8_OK ) {
168  $json = str_replace( self::$badChars, self::$badCharsEscaped, $json );
169  }
170 
171  return $json;
172  }
173 
187  public static function decode( $value, $assoc = false ) {
188  return json_decode( $value, $assoc );
189  }
190 
201  public static function parse( $value, $options = 0 ) {
202  if ( $options & self::STRIP_COMMENTS ) {
203  $value = self::stripComments( $value );
204  }
205  $assoc = ( $options & self::FORCE_ASSOC ) !== 0;
206  $result = json_decode( $value, $assoc );
207  $code = json_last_error();
208 
209  if ( $code === JSON_ERROR_SYNTAX && ( $options & self::TRY_FIXING ) !== 0 ) {
210  // The most common error is the trailing comma in a list or an object.
211  // We cannot simply replace /,\s*[}\]]/ because it could be inside a string value.
212  // But we could use the fact that JSON does not allow multi-line string values,
213  // And remove trailing commas if they are et the end of a line.
214  // JSON only allows 4 control characters: [ \t\r\n]. So we must not use '\s' for matching.
215  // Regex match ,]<any non-quote chars>\n or ,\n] with optional spaces/tabs.
216  $count = 0;
217  $value =
218  preg_replace( '/,([ \t]*[}\]][^"\r\n]*([\r\n]|$)|[ \t]*[\r\n][ \t\r\n]*[}\]])/', '$1',
219  $value, - 1, $count );
220  if ( $count > 0 ) {
221  $result = json_decode( $value, $assoc );
222  if ( JSON_ERROR_NONE === json_last_error() ) {
223  // Report warning
224  $st = Status::newGood( $result );
225  $st->warning( wfMessage( 'json-warn-trailing-comma' )->numParams( $count ) );
226  return $st;
227  }
228  }
229  }
230 
231  switch ( $code ) {
232  case JSON_ERROR_NONE:
233  return Status::newGood( $result );
234  default:
235  return Status::newFatal( wfMessage( 'json-error-unknown' )->numParams( $code ) );
236  case JSON_ERROR_DEPTH:
237  $msg = 'json-error-depth';
238  break;
239  case JSON_ERROR_STATE_MISMATCH:
240  $msg = 'json-error-state-mismatch';
241  break;
242  case JSON_ERROR_CTRL_CHAR:
243  $msg = 'json-error-ctrl-char';
244  break;
245  case JSON_ERROR_SYNTAX:
246  $msg = 'json-error-syntax';
247  break;
248  case JSON_ERROR_UTF8:
249  $msg = 'json-error-utf8';
250  break;
251  case JSON_ERROR_RECURSION:
252  $msg = 'json-error-recursion';
253  break;
254  case JSON_ERROR_INF_OR_NAN:
255  $msg = 'json-error-inf-or-nan';
256  break;
257  case JSON_ERROR_UNSUPPORTED_TYPE:
258  $msg = 'json-error-unsupported-type';
259  break;
260  }
261  return Status::newFatal( $msg );
262  }
263 
272  public static function stripComments( $json ) {
273  // Ensure we have a string
274  $str = (string) $json;
275  $buffer = '';
276  $maxLen = strlen( $str );
277  $mark = 0;
278 
279  $inString = false;
280  $inComment = false;
281  $multiline = false;
282 
283  for ( $idx = 0; $idx < $maxLen; $idx++ ) {
284  switch ( $str[$idx] ) {
285  case '"':
286  $lookBehind = ( $idx - 1 >= 0 ) ? $str[$idx - 1] : '';
287  if ( !$inComment && $lookBehind !== '\\' ) {
288  // Either started or ended a string
289  $inString = !$inString;
290  }
291  break;
292 
293  case '/':
294  $lookAhead = ( $idx + 1 < $maxLen ) ? $str[$idx + 1] : '';
295  $lookBehind = ( $idx - 1 >= 0 ) ? $str[$idx - 1] : '';
296  if ( $inString ) {
297  continue;
298 
299  } elseif ( !$inComment &&
300  ( $lookAhead === '/' || $lookAhead === '*' )
301  ) {
302  // Transition into a comment
303  // Add characters seen to buffer
304  $buffer .= substr( $str, $mark, $idx - $mark );
305  // Consume the look ahead character
306  $idx++;
307  // Track state
308  $inComment = true;
309  $multiline = $lookAhead === '*';
310 
311  } elseif ( $multiline && $lookBehind === '*' ) {
312  // Found the end of the current comment
313  $mark = $idx + 1;
314  $inComment = false;
315  $multiline = false;
316  }
317  break;
318 
319  case "\n":
320  if ( $inComment && !$multiline ) {
321  // Found the end of the current comment
322  $mark = $idx + 1;
323  $inComment = false;
324  }
325  break;
326  }
327  }
328  if ( $inComment ) {
329  // Comment ends with input
330  // Technically we should check to ensure that we aren't in
331  // a multiline comment that hasn't been properly ended, but this
332  // is a strip filter, not a validating parser.
333  $mark = $maxLen;
334  }
335  // Add final chunk to buffer before returning
336  return $buffer . substr( $str, $mark, $maxLen - $mark );
337  }
338 }
static newFatal($message)
Factory function for fatal errors.
Definition: StatusValue.php:63
const ALL_OK
Skip escaping as many characters as reasonably possible.
Definition: FormatJson.php:55
This code would result in ircNotify being run twice when an article is and once for brion Hooks can return three possible true was required This is the default since MediaWiki *some string
Definition: hooks.txt:177
$value
static $badChars
Characters problematic in JavaScript.
Definition: FormatJson.php:97
static parse($value, $options=0)
Decodes a JSON string.
Definition: FormatJson.php:201
The index of the header message $result[1]=The index of the body text message $result[2 through n]=Parameters passed to body text message.Please note the header message cannot receive/use parameters. 'ImportHandleLogItemXMLTag':When parsing a XML tag in a log item.Return false to stop further processing of the tag $reader:XMLReader object $logInfo:Array of information 'ImportHandlePageXMLTag':When parsing a XML tag in a page.Return false to stop further processing of the tag $reader:XMLReader object &$pageInfo:Array of information 'ImportHandleRevisionXMLTag':When parsing a XML tag in a page revision.Return false to stop further processing of the tag $reader:XMLReader object $pageInfo:Array of page information $revisionInfo:Array of revision information 'ImportHandleToplevelXMLTag':When parsing a top level XML tag.Return false to stop further processing of the tag $reader:XMLReader object 'ImportHandleUploadXMLTag':When parsing a XML tag in a file upload.Return false to stop further processing of the tag $reader:XMLReader object $revisionInfo:Array of information 'ImportLogInterwikiLink':Hook to change the interwiki link used in log entries and edit summaries for transwiki imports.&$fullInterwikiPrefix:Interwiki prefix, may contain colons.&$pageTitle:String that contains page title. 'ImportSources':Called when reading from the $wgImportSources configuration variable.Can be used to lazy-load the import sources list.&$importSources:The value of $wgImportSources.Modify as necessary.See the comment in DefaultSettings.php for the detail of how to structure this array. 'InfoAction':When building information to display on the action=info page.$context:IContextSource object &$pageInfo:Array of information 'InitializeArticleMaybeRedirect':MediaWiki check to see if title is a redirect.&$title:Title object for the current page &$request:WebRequest &$ignoreRedirect:boolean to skip redirect check &$target:Title/string of redirect target &$article:Article object 'InternalParseBeforeLinks':during Parser's internalParse method before links but after nowiki/noinclude/includeonly/onlyinclude and other processings.&$parser:Parser object &$text:string containing partially parsed text &$stripState:Parser's internal StripState object 'InternalParseBeforeSanitize':during Parser's internalParse method just before the parser removes unwanted/dangerous HTML tags and after nowiki/noinclude/includeonly/onlyinclude and other processings.Ideal for syntax-extensions after template/parser function execution which respect nowiki and HTML-comments.&$parser:Parser object &$text:string containing partially parsed text &$stripState:Parser's internal StripState object 'InterwikiLoadPrefix':When resolving if a given prefix is an interwiki or not.Return true without providing an interwiki to continue interwiki search.$prefix:interwiki prefix we are looking for.&$iwData:output array describing the interwiki with keys iw_url, iw_local, iw_trans and optionally iw_api and iw_wikiid. 'InvalidateEmailComplete':Called after a user's email has been invalidated successfully.$user:user(object) whose email is being invalidated 'IRCLineURL':When constructing the URL to use in an IRC notification.Callee may modify $url and $query, URL will be constructed as $url.$query &$url:URL to index.php &$query:Query string $rc:RecentChange object that triggered url generation 'IsFileCacheable':Override the result of Article::isFileCacheable()(if true) &$article:article(object) being checked 'IsTrustedProxy':Override the result of IP::isTrustedProxy() &$ip:IP being check &$result:Change this value to override the result of IP::isTrustedProxy() 'IsUploadAllowedFromUrl':Override the result of UploadFromUrl::isAllowedUrl() $url:URL used to upload from &$allowed:Boolean indicating if uploading is allowed for given URL 'isValidEmailAddr':Override the result of Sanitizer::validateEmail(), for instance to return false if the domain name doesn't match your organization.$addr:The e-mail address entered by the user &$result:Set this and return false to override the internal checks 'isValidPassword':Override the result of User::isValidPassword() $password:The password entered by the user &$result:Set this and return false to override the internal checks $user:User the password is being validated for 'Language::getMessagesFileName':$code:The language code or the language we're looking for a messages file for &$file:The messages file path, you can override this to change the location. 'LanguageGetMagic':DEPRECATED!Use $magicWords in a file listed in $wgExtensionMessagesFiles instead.Use this to define synonyms of magic words depending of the language &$magicExtensions:associative array of magic words synonyms $lang:language code(string) 'LanguageGetNamespaces':Provide custom ordering for namespaces or remove namespaces.Do not use this hook to add namespaces.Use CanonicalNamespaces for that.&$namespaces:Array of namespaces indexed by their numbers 'LanguageGetSpecialPageAliases':DEPRECATED!Use $specialPageAliases in a file listed in $wgExtensionMessagesFiles instead.Use to define aliases of special pages names depending of the language &$specialPageAliases:associative array of magic words synonyms $lang:language code(string) 'LanguageGetTranslatedLanguageNames':Provide translated language names.&$names:array of language code=> language name $code:language of the preferred translations 'LanguageLinks':Manipulate a page's language links.This is called in various places to allow extensions to define the effective language links for a page.$title:The page's Title.&$links:Associative array mapping language codes to prefixed links of the form"language:title".&$linkFlags:Associative array mapping prefixed links to arrays of flags.Currently unused, but planned to provide support for marking individual language links in the UI, e.g.for featured articles. 'LanguageSelector':Hook to change the language selector available on a page.$out:The output page.$cssClassName:CSS class name of the language selector. 'LinkBegin':DEPRECATED!Use HtmlPageLinkRendererBegin instead.Used when generating internal and interwiki links in Linker::link(), before processing starts.Return false to skip default processing and return $ret.See documentation for Linker::link() for details on the expected meanings of parameters.$skin:the Skin object $target:the Title that the link is pointing to &$html:the contents that the< a > tag should have(raw HTML) $result
Definition: hooks.txt:1934
const UTF8_OK
Skip escaping most characters above U+007F for readability and compactness.
Definition: FormatJson.php:34
static $badCharsEscaped
Escape sequences for characters listed in FormatJson::$badChars.
Definition: FormatJson.php:105
const TRY_FIXING
If set, attempts to fix invalid json.
Definition: FormatJson.php:71
either a unescaped string or a HtmlArmor object after in associative array form externallinks including delete and has completed for all link tables whether this was an auto creation default is conds Array Extra conditions for the No matching items in log is displayed if loglist is empty msgKey Array If you want a nice box with a set this to the key of the message First element is the message additional optional elements are parameters for the key that are processed with wfMessage() -> params() ->parseAsBlock()-offset Set to overwrite offset parameter in $wgRequest set to ''to unsetoffset-wrap String Wrap the message in html(usually something like"&lt
this hook is for auditing only RecentChangesLinked and Watchlist RecentChangesLinked and Watchlist e g Watchlist removed from all revisions and log entries to which it was applied This gives extensions a chance to take it off their books as the deletion has already been partly carried out by this point or something similar the user will be unable to create the tag set and then return false from the hook function Ensure you consume the ChangeTagAfterDelete hook to carry out custom deletion actions as context called by AbstractContent::getParserOutput May be used to override the normal model specific rendering of page content as context as context $options
Definition: hooks.txt:1046
const FORCE_ASSOC
If set, treat json objects '{...}' as associative arrays.
Definition: FormatJson.php:64
const STRIP_COMMENTS
If set, strip comments from input before parsing as JSON.
Definition: FormatJson.php:78
static encode($value, $pretty=false, $escaping=0)
Returns the JSON representation of a value.
Definition: FormatJson.php:127
$buffer
static newGood($value=null)
Factory function for good results.
Definition: StatusValue.php:76
const WS_CLEANUP_REGEX
Regex that matches whitespace inside empty arrays and objects.
Definition: FormatJson.php:89
this hook is for auditing only or null if authentication failed before getting that far or null if we can t even determine that probably a stub it is not rendered in wiki pages or galleries in category pages allow injecting custom HTML after the section Any uses of the hook need to handle escaping see BaseTemplate::getToolbox and BaseTemplate::makeListItem for details on the format of individual items inside of this array or by returning and letting standard HTTP rendering take place modifiable or by returning false and taking over the output modifiable & $code
Definition: hooks.txt:802
injection txt This is an overview of how MediaWiki makes use of dependency injection The design described here grew from the discussion of RFC T384 The term dependency this means that anything an object needs to operate should be injected from the the object itself should only know narrow no concrete implementation of the logic it relies on The requirement to inject everything typically results in an architecture that based on two main types of and essentially stateless service objects that use other service objects to operate on the value objects As of the beginning MediaWiki is only starting to use the DI approach Much of the code still relies on global state or direct resulting in a highly cyclical dependency which acts as the top level factory for services in MediaWiki which can be used to gain access to default instances of various services MediaWikiServices however also allows new services to be defined and default services to be redefined Services are defined or redefined by providing a callback the instantiator that will return a new instance of the service When it will create an instance of MediaWikiServices and populate it with the services defined in the files listed by thereby bootstrapping the DI framework Per $wgServiceWiringFiles lists includes ServiceWiring php
Definition: injection.txt:35
$count
static decode($value, $assoc=false)
Decodes a JSON string.
Definition: FormatJson.php:187
static stripComments($json)
Remove multiline and single line comments from an otherwise valid JSON input string.
Definition: FormatJson.php:272
const XMLMETA_OK
Skip escaping the characters '<', '>', and '&', which have special meanings in HTML and XML...
Definition: FormatJson.php:46